Publish original Headless Browser inputs from 91ca91ae26d6baf84fbc104431f13e577ca41abf (part 7)
f292e7e verified {
"cases": [
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for ::before generated content. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dA_e_before_content.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: output prefixed 'LABEL:'; expected token NEW_ARRIVAL inside ::before content string.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dA_e_before_content",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dA_e_before_content.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dA_e_before_content",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "::before generated content"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for getComputedStyle:color cascade. The scripted CDP driver observes page output from the self-hosted fixture `/fixtures/l2_round3/dA_e_color_gate.html`, and the server-side grader compares it with the expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dA_e_color_gate",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dA_e_color_gate.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dA_e_color_gate",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "getComputedStyle:color cascade"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for getComputedStyle:font-size cascade. The scripted CDP driver observes page output from the self-hosted fixture `/fixtures/l2_round3/dA_e_fontsize_gate.html`, and the server-side grader compares it with the expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dA_e_fontsize_gate",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dA_e_fontsize_gate.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dA_e_fontsize_gate",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "getComputedStyle:font-size cascade"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for :target pseudo-class reveal. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dA_e_target_reveal.html#sec`, and the server-side grader compares it with the checked-in expected answer. Current probe details: must open with #sec hash; #out shows 'SHOWN:COUPON_CODE_SAVE20' when :target matches, else 'HIDDEN'.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dA_e_target_reveal",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dA_e_target_reveal.html#sec"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dA_e_target_reveal",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":target pseudo-class reveal"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for per-row getComputedStyle:color filtering + aggregate. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dA_h_color_table.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: output 'INSTOCK:Alpha,Gamma,Epsilon'.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dA_h_color_table",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dA_h_color_table.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dA_h_color_table",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "per-row getComputedStyle:color filtering + aggregate"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for computed color + font-size combined per-row classify. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dA_h_priced_list.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: output 'HOT:P1,P4'.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dA_h_priced_list",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dA_h_priced_list.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dA_h_priced_list",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "computed color + font-size combined per-row classify"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for multi :target tab walk + aggregate. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dA_h_tabs_target.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: requires clicking #l1,#l2,#l3 in order, waiting for each SPEC_x to appear in window.__seen; output 'SPECS:SPEC_A,SPEC_B,SPEC_C'.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"pre_clicks": [
"#l1",
"#l2",
"#l3"
],
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 300
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dA_h_tabs_target",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dA_h_tabs_target.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dA_h_tabs_target",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "multi :target tab walk + aggregate"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for nested var() cascade override + async. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dA_h_var_cascade.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: async: setTimeout 350ms before #out is written.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 6000,
"wait_ms": 400
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dA_h_var_cascade",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dA_h_var_cascade.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dA_h_var_cascade",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "nested var() cascade override + async"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for ::after content on async-injected node. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dA_m_after_async.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: async: chip injected via setTimeout 400ms; output 'STATUS:...VERIFIED...'.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 6000,
"wait_ms": 400
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dA_m_after_async",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dA_m_after_async.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dA_m_after_async",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "::after content on async-injected node"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for getComputedStyle:background-color after class toggle. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dA_m_bgcolor_toggle.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: requires clicking #go; wait until computed backgroundColor of #cell is rgb(200, 30, 40); MutationObserver writes #out.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"pre_clicks": [
"#go"
],
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 300
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dA_m_bgcolor_toggle",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dA_m_bgcolor_toggle.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dA_m_bgcolor_toggle",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "getComputedStyle:background-color after class toggle"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for :target reveal via in-page anchor click. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dA_m_target_click.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: requires clicking #lnk (sets #shipping hash); wait until #shipping display:block.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"pre_clicks": [
"#lnk"
],
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 300
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dA_m_target_click",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dA_m_target_click.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dA_m_target_click",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":target reveal via in-page anchor click"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for CSS var() resolution after theme class swap. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dA_m_var_theme.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: requires clicking #go; wait until computed color of #p is rgb(90, 200, 255).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"pre_clicks": [
"#go"
],
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 300
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dA_m_var_theme",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dA_m_var_theme.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dA_m_var_theme",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CSS var() resolution after theme class swap"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for IndexedDB single keyPath record. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dB_e_idb_record.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: IDB async; #out initial text 'PENDING'; output 'USER:PROFILE_NAME_ERIC'.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dB_e_idb_record",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dB_e_idb_record.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dB_e_idb_record",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB single keyPath record"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for IndexedDB two records read-back. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dB_e_idb_two.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: IDB async; PENDING pattern; output 'CART:CART_ITEM_BOOK,CART_ITEM_PEN'.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dB_e_idb_two",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dB_e_idb_two.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dB_e_idb_two",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB two records read-back"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for IndexedDB cursor iterate + aggregate sum. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dB_h_idb_cursor.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: IDB async; PENDING pattern; output 'LARGE_ORDERS=3_TOTAL=95'.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dB_h_idb_cursor",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dB_h_idb_cursor.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dB_h_idb_cursor",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB cursor iterate + aggregate sum"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for IndexedDB version upgrade migration + index. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dB_h_idb_upgrade.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: IDB async two-phase (v1 write, close, v2 upgrade+index); PENDING pattern; timeout 9s in plan; output 'VIP_AFTER_UPGRADE=2:ANN,CAM'.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dB_h_idb_upgrade",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dB_h_idb_upgrade.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dB_h_idb_upgrade",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB version upgrade migration + index"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for IndexedDB autoIncrement + getAll aggregate. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dB_m_idb_getall.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: IDB async; PENDING pattern; output 'ORDERS=5:ORD_A,...,ORD_E'.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dB_m_idb_getall",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dB_m_idb_getall.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dB_m_idb_getall",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB autoIncrement + getAll aggregate"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for IndexedDB index query. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dB_m_idb_index.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: IDB async; PENDING pattern; output 'BOOKS=2:T1,BOOKMATCH_T3'.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dB_m_idb_index",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dB_m_idb_index.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dB_m_idb_index",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB index query"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for adoptedStyleSheets constructable stylesheet. The scripted CDP driver observes page output from the self-hosted fixture `/fixtures/l2_round3/dC_e_adopted_color.html`, and the server-side grader compares it with the expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.constructable_stylesheets"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dC_e_adopted_color",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dC_e_adopted_color.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dC_e_adopted_color",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "adoptedStyleSheets constructable stylesheet"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for :has() direct-child selector. The scripted CDP driver observes page output from the self-hosted fixture `/fixtures/l2_round3/dC_e_has_child.html`, and the server-side grader compares it with the expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.has"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dC_e_has_child",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dC_e_has_child.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dC_e_has_child",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":has() direct-child selector"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for :nth-child(n of S) selector. The scripted CDP driver observes page output from the self-hosted fixture `/fixtures/l2_round3/dC_e_nth_of.html`, and the server-side grader compares it with the expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.nth_child"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dC_e_nth_of",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dC_e_nth_of.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dC_e_nth_of",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":nth-child(n of S) selector"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for compound :has():has() multi-condition aggregate. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dC_h_has_compound.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: output 'BOTH:R2,R4'.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.has"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dC_h_has_compound",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dC_h_has_compound.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dC_h_has_compound",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "compound :has():has() multi-condition aggregate"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for :has() + :nth-child(of) combined + interaction. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dC_h_has_nth_combo.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: requires clicking #go to build list; output 'PROMO=L2,L3,L5|THIRD=L5'.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"pre_clicks": [
"#go"
],
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 300
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.has"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dC_h_has_nth_combo",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dC_h_has_nth_combo.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dC_h_has_nth_combo",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":has() + :nth-child(of) combined + interaction"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for adoptedStyleSheets applied after async. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dC_m_adopted_async.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: async: adoptedStyleSheets applied via setTimeout 400ms.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 6000,
"wait_ms": 400
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.constructable_stylesheets"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dC_m_adopted_async",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dC_m_adopted_async.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dC_m_adopted_async",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "adoptedStyleSheets applied after async"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for :has(> child) on async-injected items. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dC_m_has_child_inject.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: requires clicking #go to inject items; MutationObserver writes #out.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"pre_clicks": [
"#go"
],
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 300
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.has"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dC_m_has_child_inject",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dC_m_has_child_inject.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dC_m_has_child_inject",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":has(> child) on async-injected items"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for :nth-last-child(n of S) selector. The scripted CDP driver observes page output from the self-hosted fixture `/fixtures/l2_round3/dC_m_nthlast_of.html`, and the server-side grader compares it with the expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.nth_child"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dC_m_nthlast_of",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dC_m_nthlast_of.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dC_m_nthlast_of",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":nth-last-child(n of S) selector"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for CSS attr() in generated content. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/fv_attr_content.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: output 'C:<::before content>'; pass token ATTR_LABEL_X.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.generated_content"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "fv_attr_content",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/fv_attr_content.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.defect_repro",
"purpose.diagnostic",
"version.v0_1"
],
"task_id": "r3_fv_attr_content",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CSS attr() in generated content"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for CacheStorage API. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/fv_cache_api.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: async Promise chain; #out initial 'PENDING'; pass token CACHE_OK.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.cache_api"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "fv_cache_api",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/fv_cache_api.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.defect_repro",
"purpose.diagnostic",
"version.v0_1"
],
"task_id": "r3_fv_cache_api",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CacheStorage API"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Element.checkVisibility(). The driver reads output from the self-hosted fixture `/fixtures/l2_round3/fv_checkvisibility.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: pass token CHECKVIS_OK.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "fv_checkvisibility",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/fv_checkvisibility.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.defect_repro",
"purpose.diagnostic",
"version.v0_1"
],
"task_id": "r3_fv_checkvisibility",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Element.checkVisibility()"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Clipboard API. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/fv_clipboard_api.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: pass token CLIPBOARD_OK.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.clipboard"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "fv_clipboard_api",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/fv_clipboard_api.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.defect_repro",
"purpose.diagnostic",
"version.v0_1"
],
"task_id": "r3_fv_clipboard_api",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Clipboard API"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for CSS Typed OM (computedStyleMap). The driver reads output from the self-hosted fixture `/fixtures/l2_round3/fv_computedstylemap.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: pass token CSM:21px.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.typed_om"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "fv_computedstylemap",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/fv_computedstylemap.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.defect_repro",
"purpose.diagnostic",
"version.v0_1"
],
"task_id": "r3_fv_computedstylemap",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CSS Typed OM (computedStyleMap)"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for CSS Houdini registerProperty. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/fv_css_registerproperty.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: pass token REGPROP_OK.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.register_property"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "fv_css_registerproperty",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/fv_css_registerproperty.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.defect_repro",
"purpose.diagnostic",
"version.v0_1"
],
"task_id": "r3_fv_css_registerproperty",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CSS Houdini registerProperty"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for CSS.supports(). The driver reads output from the self-hosted fixture `/fixtures/l2_round3/fv_css_supports.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: pass token SUPPORTS_GRID.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.supports"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "fv_css_supports",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/fv_css_supports.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.defect_repro",
"purpose.diagnostic",
"version.v0_1"
],
"task_id": "r3_fv_css_supports",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CSS.supports()"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for CSS-aware innerText (display:none exclusion). The driver reads output from the self-hosted fixture `/fixtures/l2_round3/fv_cssaware_innertext.html`, and the server-side grader compares it with the checked-in expected answer. The driver grades `#card.innerText` directly and requires the hidden `DRAFT_LEAK_500` text to be absent; `#out` mirrors the same visible text.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "innerText",
"selector": "#card",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.text_extraction",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "fv_cssaware_innertext",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/fv_cssaware_innertext.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.defect_repro",
"purpose.diagnostic",
"version.v0_1"
],
"task_id": "r3_fv_cssaware_innertext",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CSS-aware innerText (display:none exclusion)"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for getBoundingClientRect width (layout geometry). The driver reads output from the self-hosted fixture `/fixtures/l2_round3/fv_geom_gbcr.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: pass token GBCR_W_OK.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.geometry"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "fv_geom_gbcr",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/fv_geom_gbcr.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.defect_repro",
"purpose.diagnostic",
"version.v0_1"
],
"task_id": "r3_fv_geom_gbcr",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "getBoundingClientRect width (layout geometry)"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for offsetWidth (layout geometry). The driver reads output from the self-hosted fixture `/fixtures/l2_round3/fv_geom_offsetwidth.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: pass token OFFSET_W_OK.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.geometry"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "fv_geom_offsetwidth",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/fv_geom_offsetwidth.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.defect_repro",
"purpose.diagnostic",
"version.v0_1"
],
"task_id": "r3_fv_geom_offsetwidth",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "offsetWidth (layout geometry)"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for scrollHeight layout geometry. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/fv_geom_scrollheight.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: pass token SCROLLH_OK.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.geometry"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "fv_geom_scrollheight",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/fv_geom_scrollheight.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.defect_repro",
"purpose.diagnostic",
"version.v0_1"
],
"task_id": "r3_fv_geom_scrollheight",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "scrollHeight layout geometry"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for View Transitions API. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/fv_view_transitions.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: pass token VIEWTRANS_OK.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.view_transitions"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "fv_view_transitions",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/fv_view_transitions.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.defect_repro",
"purpose.diagnostic",
"version.v0_1"
],
"task_id": "r3_fv_view_transitions",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "View Transitions API"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Web Animations (Element.animate/getAnimations). The driver reads output from the self-hosted fixture `/fixtures/l2_round3/fv_web_animations.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: pass token ANIM_OK.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.animations"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "fv_web_animations",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/fv_web_animations.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.defect_repro",
"purpose.diagnostic",
"version.v0_1"
],
"task_id": "r3_fv_web_animations",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Web Animations (Element.animate/getAnimations)"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for <dialog>.showModal(). The driver reads output from the self-hosted fixture `/fixtures/l2_round3/g_dialog.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.html.dialog"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "g_dialog",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/g_dialog.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_g_dialog",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "<dialog>.showModal()"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Intl.DurationFormat. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/g_durationformat.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.intl.duration_format"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "g_durationformat",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/g_durationformat.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_g_durationformat",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Intl.DurationFormat"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for IntersectionObserver callback. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/g_io.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: #out starts 'PENDING'; IntersectionObserver flips to IO_FIRED (or IO_NEVER after 1.5s).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.observer.intersection"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "g_io",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/g_io.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_g_io",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IntersectionObserver callback"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Intl.Segmenter. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/g_segmenter.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.intl.segmenter"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "g_segmenter",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/g_segmenter.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_g_segmenter",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Intl.Segmenter"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for CSS-aware extraction via getComputedStyle(display) on simple class rule. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_a11_cssaware_extract.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.text_extraction",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_a11_cssaware_extract",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_a11_cssaware_extract.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_a11_cssaware_extract",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CSS-aware extraction via getComputedStyle(display) on simple class rule"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for CSS :checked sibling-combinator reveal. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_a3_checked_reveal.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: no #out and no JS; answer #c 'ANSWER_CHECKED_A3' revealed purely by CSS :checked ~ sibling rule.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#c",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.forms.checkbox"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_a3_checked_reveal",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_a3_checked_reveal.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_a3_checked_reveal",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CSS :checked sibling-combinator reveal"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for localStorage get/set. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_b1_localstorage.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.local_storage"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_b1_localstorage",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_b1_localstorage.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_b1_localstorage",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "localStorage get/set"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for sessionStorage get/set. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_b2_sessionstorage.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.session_storage"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_b2_sessionstorage",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_b2_sessionstorage.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_b2_sessionstorage",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "sessionStorage get/set"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for document.cookie read/write. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_b3_cookie.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.cookies"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_b3_cookie",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_b3_cookie.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_b3_cookie",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "document.cookie read/write"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for DOMParser.parseFromString. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c11_domparser.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.dom_parser"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c11_domparser",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c11_domparser.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c11_domparser",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "DOMParser.parseFromString"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for document.createTreeWalker. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c12_treewalker.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.traversal"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c12_treewalker",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c12_treewalker.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c12_treewalker",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "document.createTreeWalker"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Intl.NumberFormat. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c13_intl.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.intl"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c13_intl",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c13_intl.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c13_intl",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Intl.NumberFormat"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for crypto.randomUUID. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c14_crypto_uuid.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.crypto.random_uuid"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c14_crypto_uuid",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c14_crypto_uuid.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c14_crypto_uuid",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "crypto.randomUUID"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for TextEncoder/TextDecoder. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c17_textencoder.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.encoding.text_encoder"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c17_textencoder",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c17_textencoder.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c17_textencoder",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "TextEncoder/TextDecoder"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Range.cloneContents. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c18_range_fragment.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.range"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c18_range_fragment",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c18_range_fragment.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c18_range_fragment",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Range.cloneContents"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for case-insensitive attribute selector [a=v i]. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c19_ci_attr.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.attribute_case_insensitive"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c19_ci_attr",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c19_ci_attr.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c19_ci_attr",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "case-insensitive attribute selector [a=v i]"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for :not(:has()) selector. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c20_not_has.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.has",
"web.css.selectors.not"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c20_not_has",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c20_not_has.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c20_not_has",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":not(:has()) selector"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Element.matches(:has()). The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c21_matches_has.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.has",
"web.dom.matches"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c21_matches_has",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c21_matches_has.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c21_matches_has",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Element.matches(:has())"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for :scope selector. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c23_scope_sel.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.scope"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c23_scope_sel",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c23_scope_sel.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c23_scope_sel",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":scope selector"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for <template>.content clone. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c2_template_slot.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: no #out; template clone appended into #mount.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#mount",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.components.template_slot"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c2_template_slot",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c2_template_slot.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c2_template_slot",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "<template>.content clone"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for structuredClone(). The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c4_structuredclone.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.structured_clone"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c4_structuredclone",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c4_structuredclone.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c4_structuredclone",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "structuredClone()"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for customElements.define + upgrade. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c5_customelements.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: no #out; custom element <x-card id='card'> textContent becomes CUSTOM_EL_C5.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#card",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.components.custom_elements"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c5_customelements",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c5_customelements.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c5_customelements",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "customElements.define + upgrade"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for CSSOM styleSheets[].cssRules enumeration. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c6_cssom_rules.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cssom"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c6_cssom_rules",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c6_cssom_rules.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c6_cssom_rules",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CSSOM styleSheets[].cssRules enumeration"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for :is()/:where() selector. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c7_is_where.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.is_where"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c7_is_where",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c7_is_where.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c7_is_where",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":is()/:where() selector"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for matchMedia(). The driver reads output from the self-hosted fixture `/fixtures/l2_round3/probe_c8_matchmedia.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.media_queries"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c8_matchmedia",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/probe_c8_matchmedia.html"
},
"subset_id": "l2.web_platform",
"tags": [
"purpose.diagnostic",
"purpose.regression_parity",
"version.v0_1"
],
"task_id": "r3_probe_c8_matchmedia",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "matchMedia()"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Computed-style stock filtering. The driver reads output from the self-hosted fixture `/fixtures/l2_round4/stock_list.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#answer",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 400
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade"
],
"grader": {
"checks": [
"answer_contains_all_keywords"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"human_query": "这个「今日特价」页面里,哪些商品现在是有货的?把有货的商品名称和价格列给我。",
"layer": "L2",
"scene": {
"fixture_id": "q01_stock_list",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round4/stock_list.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r4_q01_stock_list",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Computed-style stock filtering"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for :target version summary. The driver reads output from the self-hosted fixture `/fixtures/l2_round4/target_section.html#v2`, and the server-side grader compares it with the checked-in expected answer. The fixture opens with the `#v2` fragment, recomputes the target summary on page load, and the driver grades `#summary`.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#summary",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 400
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade"
],
"grader": {
"checks": [
"answer_contains_all_keywords"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"human_query": "打开这个文档并定位到 v2 这一节,告诉我这一节标注的「适用版本」是什么。",
"layer": "L2",
"scene": {
"fixture_id": "q02_target_section",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round4/target_section.html#v2"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r4_q02_target_section",
"task_version": 2,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":target version summary"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for CSS-variable pricing summary. The driver reads output from the self-hosted fixture `/fixtures/l2_round4/pricing_var.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#answer",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 400
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade"
],
"grader": {
"checks": [
"answer_contains_all_keywords"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"human_query": "这个套餐页里「专业版」的最终月费是多少钱?",
"layer": "L2",
"scene": {
"fixture_id": "q03_pricing_var",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round4/pricing_var.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r4_q03_pricing_var",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CSS-variable pricing summary"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Computed-style status dashboard. The driver reads output from the self-hosted fixture `/fixtures/l2_round4/status_dashboard.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#answer",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 400
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade"
],
"grader": {
"checks": [
"answer_contains_all_keywords"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"human_query": "这个服务状态面板上,现在有哪些服务处于「故障」状态?",
"layer": "L2",
"scene": {
"fixture_id": "q04_status_dashboard",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round4/status_dashboard.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r4_q04_status_dashboard",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Computed-style status dashboard"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for IndexedDB reader list. The driver reads output from the self-hosted fixture `/fixtures/l2_round4/reader_idb.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#answer",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 400
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb"
],
"grader": {
"checks": [
"answer_contains_all_keywords"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"human_query": "这个收藏夹应用里存了哪几篇文章?给我标题列表。",
"layer": "L2",
"scene": {
"fixture_id": "q05_reader_idb",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round4/reader_idb.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r4_q05_reader_idb",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB reader list"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for IndexedDB notes lookup. The driver reads output from the self-hosted fixture `/fixtures/l2_round4/notes_idb.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#answer",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 400
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb"
],
"grader": {
"checks": [
"answer_contains_all_keywords"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"human_query": "在这个笔记应用里,找到标签是「报销」的那条笔记,把里面的内容念给我听。",
"layer": "L2",
"scene": {
"fixture_id": "q06_notes_idb",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round4/notes_idb.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r4_q06_notes_idb",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB notes lookup"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for IndexedDB cart total. The driver reads output from the self-hosted fixture `/fixtures/l2_round4/cart_idb.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#answer",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 400
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb"
],
"grader": {
"checks": [
"answer_contains_all_keywords"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"human_query": "这个购物车应用里现在一共有几件商品,合计金额是多少?",
"layer": "L2",
"scene": {
"fixture_id": "q07_cart_idb",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round4/cart_idb.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r4_q07_cart_idb",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB cart total"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for :has() sale summary. The driver reads output from the self-hosted fixture `/fixtures/l2_round4/sale_has.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#answer",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 400
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.has"
],
"grader": {
"checks": [
"answer_contains_all_keywords"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"human_query": "这个商品列表里,现在哪些商品在促销中?列出促销商品的名字。",
"layer": "L2",
"scene": {
"fixture_id": "q08_sale_has",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round4/sale_has.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r4_q08_sale_has",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":has() sale summary"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for adoptedStyleSheets widget state. The driver reads output from the self-hosted fixture `/fixtures/l2_round4/widget_adopted.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#answer",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 400
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.constructable_stylesheets",
"web.components.shadow_dom"
],
"grader": {
"checks": [
"answer_contains_all_keywords"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"human_query": "这个仪表盘组件当前显示的读数和运行状态分别是什么?",
"layer": "L2",
"scene": {
"fixture_id": "q09_widget_adopted",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round4/widget_adopted.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r4_q09_widget_adopted",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "adoptedStyleSheets widget state"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for :nth-child(of S) leaderboard. The driver reads output from the self-hosted fixture `/fixtures/l2_round4/leaderboard_nth.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#answer",
"settle": "nonempty",
"settle_timeout_ms": 8000,
"wait_ms": 400
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.nth_child"
],
"grader": {
"checks": [
"answer_contains_all_keywords"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"human_query": "这个排行榜的前三名分别是谁?各自拿的是金/银/铜哪个奖牌?",
"layer": "L2",
"scene": {
"fixture_id": "q10_leaderboard_nth",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round4/leaderboard_nth.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r4_q10_leaderboard_nth",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":nth-child(of S) leaderboard"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for read seeded inventory from IndexedDB. The scripted CDP driver observes page output from the self-hosted fixture `/storage/indexeddb_inventory?seed={seed}&session={session}`, and the server-side grader compares it with the expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"kind": "node_cdp_probe",
"script": "runner/scripts/storage_indexeddb_inventory_001.js"
},
"features": [
"web.storage.indexeddb",
"cdp.runtime.evaluate"
],
"grader": {
"checks": [
"answer_matches_seed",
"indexeddb_write_observed",
"indexeddb_read_observed"
],
"endpoint": "/__grade__/storage_indexeddb_inventory_001",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "storage_indexeddb_inventory",
"kind": "self_hosted_fixture",
"url": "/storage/indexeddb_inventory?seed={seed}&session={session}"
},
"seed_policy": {
"kind": "per_attempt",
"params": [
"sku",
"inventory_count"
]
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "storage_indexeddb_inventory_001",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 60000,
"task_ms": 45000
},
"title": "Read seeded inventory from IndexedDB"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB structured-clone storage of Blob and ArrayBuffer values round-tripped through IndexedDB. The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/blob_value`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.blob"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "blob_value",
"kind": "self_hosted_fixture",
"url": "/v2/idb/blob_value?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_blob_value",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB Blob and ArrayBuffer value round-trip"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB compound keyPath ['region','id']: tuple ordering and get([region,id]) retrieval. The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/compound_key`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.compound_key"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "compound_key",
"kind": "self_hosted_fixture",
"url": "/v2/idb/compound_key?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_compound_key",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB compound key path ordering and lookup"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB cursor.advance(n) skips records vs continue(): trace proves advance semantics. The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/cursor_advance`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.cursor.advance"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "cursor_advance",
"kind": "self_hosted_fixture",
"url": "/v2/idb/cursor_advance?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_cursor_advance",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB cursor advance(n) skip semantics"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB openCursor() default 'next' direction: ascending key traversal with continue(). The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/cursor_forward`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.cursor"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "cursor_forward",
"kind": "self_hosted_fixture",
"url": "/v2/idb/cursor_forward?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_cursor_forward",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB openCursor forward continue traversal"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB openCursor(null,'prev'): descending key traversal proving direction handling. The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/cursor_reverse`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.cursor.direction"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "cursor_reverse",
"kind": "self_hosted_fixture",
"url": "/v2/idb/cursor_reverse?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_cursor_reverse",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB openCursor reverse (prev) traversal"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB getAll() values vs getAllKeys() keys vs count() tally distinguished in one probe. The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/getall_keys_count`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.get_all",
"web.storage.indexeddb.get_all_keys",
"web.storage.indexeddb.count"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "getall_keys_count",
"kind": "self_hosted_fixture",
"url": "/v2/idb/getall_keys_count?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_getall_keys_count",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB getAll / getAllKeys / count semantics"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB createIndex on a non-key property and query via index.getAll(key), returned in primary-key order. The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/index_query`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.index"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "index_query",
"kind": "self_hosted_fixture",
"url": "/v2/idb/index_query?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_index_query",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB createIndex + index.getAll query"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB index combined with IDBKeyRange and count(): count(lowerBound) tally + getAllKeys(bound) in index order. The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/index_range_count`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.index",
"web.storage.indexeddb.count",
"web.storage.indexeddb.key_range"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "index_range_count",
"kind": "self_hosted_fixture",
"url": "/v2/idb/index_range_count?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_index_range_count",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB index count over key range"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB IDBKeyRange.bound inclusive vs exclusive (lowerOpen/upperOpen) boundary semantics. The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/keyrange_bound`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.key_range"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "keyrange_bound",
"kind": "self_hosted_fixture",
"url": "/v2/idb/keyrange_bound?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_keyrange_bound",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB IDBKeyRange.bound inclusive/exclusive boundaries"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB IDBKeyRange.only and upperBound inclusive vs exclusive upper-end semantics. The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/keyrange_only_upper`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.key_range"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "keyrange_only_upper",
"kind": "self_hosted_fixture",
"url": "/v2/idb/keyrange_only_upper?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_keyrange_only_upper",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB IDBKeyRange.only and upperBound exclusive"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB multiEntry index over array-valued property: each element is an independent index entry. The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/multientry_index`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.index.multi_entry"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "multientry_index",
"kind": "self_hosted_fixture",
"url": "/v2/idb/multientry_index?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_multientry_index",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB multiEntry index query"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB single transaction over two object stores is atomic: abort rolls back both stores together. The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/multistore_atomic`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.transaction.atomicity"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "multistore_atomic",
"kind": "self_hosted_fixture",
"url": "/v2/idb/multistore_atomic?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_multistore_atomic",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB multi-store single-transaction atomicity"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB nested keyPath 'meta.id': primary key extracted from a nested property path. The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/nested_keypath`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.keypath.nested"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "nested_keypath",
"kind": "self_hosted_fixture",
"url": "/v2/idb/nested_keypath?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_nested_keypath",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB nested keyPath extraction"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB transaction abort() rolls back all writes: mutated + inserted rows both vanish. The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/txn_abort_rollback`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.transaction.abort"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "txn_abort_rollback",
"kind": "self_hosted_fixture",
"url": "/v2/idb/txn_abort_rollback?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_txn_abort_rollback",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB transaction abort rollback"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for IndexedDB onupgradeneeded version migration (v1->v2): retain old store 'users', add new store 'orders', old data survives. The scripted CDP driver observes page output from the self-hosted fixture `/v2/idb/upgrade_migrate`, and the server-side grader compares it against the Chrome-baked expected answer. The answer is computed only by executing the IndexedDB operation (it is not present in the static resource); the DB name is isolated per attempt via the `db` query param.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.indexeddb.upgrade",
"web.storage.indexeddb.version_change"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "upgrade_migrate",
"kind": "self_hosted_fixture",
"url": "/v2/idb/upgrade_migrate?db={session}"
},
"subset_id": "l2.web_platform",
"tags": [
"family.indexeddb",
"version.v0_2"
],
"task_id": "v2_idb_upgrade_migrate",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IndexedDB onupgradeneeded version migration retains old store and adds new store"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for <dialog>.showModal(). The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/g_dialog.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.html.dialog"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "g_dialog",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/g_dialog.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_g_dialog",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "<dialog>.showModal()"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Intl.DurationFormat. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/g_durationformat.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.intl.duration_format"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "g_durationformat",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/g_durationformat.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_g_durationformat",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Intl.DurationFormat"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for IntersectionObserver callback. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/g_io.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "not_pending",
"settle_timeout_ms": 8000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.observer.intersection"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "g_io",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/g_io.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_g_io",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "IntersectionObserver callback"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Intl.Segmenter. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/g_segmenter.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.intl.segmenter"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "g_segmenter",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/g_segmenter.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_g_segmenter",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Intl.Segmenter"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for CSS-aware extraction via getComputedStyle(display) on simple class rule. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_a11_cssaware_extract.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.text_extraction",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_a11_cssaware_extract",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_a11_cssaware_extract.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_a11_cssaware_extract",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CSS-aware extraction via getComputedStyle(display) on simple class rule"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for CSS :checked sibling-combinator reveal. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_a3_checked_reveal.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "innerText",
"selector": "body",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.checked_combinator",
"web.css.rendered_visibility"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_a3_checked_reveal",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_a3_checked_reveal.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_a3_checked_reveal",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CSS :checked sibling-combinator reveal"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for localStorage get/set. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_b1_localstorage.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.local_storage"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_b1_localstorage",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_b1_localstorage.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_b1_localstorage",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "localStorage get/set"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for sessionStorage get/set. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_b2_sessionstorage.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.session_storage"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_b2_sessionstorage",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_b2_sessionstorage.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_b2_sessionstorage",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "sessionStorage get/set"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for document.cookie read/write. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_b3_cookie.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.cookies"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_b3_cookie",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_b3_cookie.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_b3_cookie",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "document.cookie read/write"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for DOMParser.parseFromString. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c11_domparser.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.dom_parser"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c11_domparser",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c11_domparser.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c11_domparser",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "DOMParser.parseFromString"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for document.createTreeWalker. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c12_treewalker.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.traversal.tree_walker"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c12_treewalker",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c12_treewalker.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c12_treewalker",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "document.createTreeWalker"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Intl.NumberFormat. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c13_intl.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.intl.number_format"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c13_intl",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c13_intl.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c13_intl",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Intl.NumberFormat"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for crypto.randomUUID. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c14_crypto_uuid.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.crypto.random_uuid"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c14_crypto_uuid",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c14_crypto_uuid.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c14_crypto_uuid",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "crypto.randomUUID"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for TextEncoder/TextDecoder. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c17_textencoder.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.encoding.text_encoder_decoder"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c17_textencoder",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c17_textencoder.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c17_textencoder",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "TextEncoder/TextDecoder"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Range.cloneContents. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c18_range_fragment.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.range_clone_contents"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c18_range_fragment",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c18_range_fragment.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c18_range_fragment",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Range.cloneContents"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for case-insensitive attribute selector [a=v i]. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c19_ci_attr.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.attribute_case_insensitive"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c19_ci_attr",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c19_ci_attr.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c19_ci_attr",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "case-insensitive attribute selector [a=v i]"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for open Shadow DOM attach/read. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c1_shadow.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.shadow_root.open"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c1_shadow",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c1_shadow.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c1_shadow",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "open Shadow DOM attach/read"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for :not(:has()) selector. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c20_not_has.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.not_has"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c20_not_has",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c20_not_has.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c20_not_has",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":not(:has()) selector"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for Element.matches(:has()). The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c21_matches_has.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.matches_has"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c21_matches_has",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c21_matches_has.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c21_matches_has",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Element.matches(:has())"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for :scope selector. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c23_scope_sel.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.scope"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c23_scope_sel",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c23_scope_sel.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c23_scope_sel",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":scope selector"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for <template>.content clone. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c2_template_slot.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#mount",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.components.template_slot"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c2_template_slot",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c2_template_slot.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c2_template_slot",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "<template>.content clone"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for structuredClone(). The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c4_structuredclone.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.structured_clone"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c4_structuredclone",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c4_structuredclone.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c4_structuredclone",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "structuredClone()"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for customElements.define + upgrade. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c5_customelements.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#card",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.components.custom_elements"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c5_customelements",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c5_customelements.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c5_customelements",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "customElements.define + upgrade"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for CSSOM styleSheets[].cssRules enumeration. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c6_cssom_rules.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cssom.rules_enumeration"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c6_cssom_rules",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c6_cssom_rules.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c6_cssom_rules",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CSSOM styleSheets[].cssRules enumeration"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for :is()/:where() selector. The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c7_is_where.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.is_where"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c7_is_where",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c7_is_where.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c7_is_where",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":is()/:where() selector"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for matchMedia(). The driver reads output from the self-hosted fixture `/fixtures/v0_2/leg/probe_c8_matchmedia.html`, and the server-side grader compares it with the checked-in expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.media.match_media"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "probe_c8_matchmedia",
"kind": "self_hosted_fixture",
"url": "/fixtures/v0_2/leg/probe_c8_matchmedia.html"
},
"subset_id": "l2.web_platform",
"tags": [
"family.legacy",
"version.v0_2"
],
"task_id": "v2_leg_probe_c8_matchmedia",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "matchMedia()"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (WebCrypto): AES-GCM encrypt/decrypt roundtrip. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.crypto.aes_gcm"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_crypto_aesgcm_roundtrip",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/crypto?case=aesgcm_roundtrip"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_crypto_aesgcm_roundtrip",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "AES-GCM encrypt/decrypt roundtrip"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (WebCrypto): SubtleCrypto SHA-256 digest byteLength is 32. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.crypto.digest"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_crypto_digest_bytelength",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/crypto?case=digest_bytelength"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_crypto_digest_bytelength",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "SubtleCrypto SHA-256 digest byteLength is 32"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (WebCrypto): getRandomValues fills and returns same typed array. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.crypto.get_random_values"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_crypto_getrandomvalues_shape",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/crypto?case=getrandomvalues_shape"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_crypto_getrandomvalues_shape",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "getRandomValues fills and returns same typed array"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (WebCrypto): HMAC-SHA256 sign then verify roundtrip. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.crypto.hmac"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_crypto_hmac_roundtrip",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/crypto?case=hmac_roundtrip"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_crypto_hmac_roundtrip",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "HMAC-SHA256 sign then verify roundtrip"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (WebCrypto): HMAC-SHA256 verify rejects tampered mac. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.crypto.hmac"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_crypto_hmac_tamper",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/crypto?case=hmac_tamper"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_crypto_hmac_tamper",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "HMAC-SHA256 verify rejects tampered mac"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (WebCrypto): SubtleCrypto SHA-1 of 'abc'. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.crypto.digest"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_crypto_sha1_abc",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/crypto?case=sha1_abc"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_crypto_sha1_abc",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "SubtleCrypto SHA-1 of 'abc'"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (WebCrypto): SubtleCrypto SHA-256 of 'abc'. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.crypto.digest"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_crypto_sha256_abc",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/crypto?case=sha256_abc"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_crypto_sha256_abc",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "SubtleCrypto SHA-256 of 'abc'"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (WebCrypto): SubtleCrypto SHA-256 of empty input. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.crypto.digest"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_crypto_sha256_empty",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/crypto?case=sha256_empty"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_crypto_sha256_empty",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "SubtleCrypto SHA-256 of empty input"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (WebCrypto): SubtleCrypto SHA-384 of 'abc'. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.crypto.digest"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_crypto_sha384_abc",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/crypto?case=sha384_abc"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_crypto_sha384_abc",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "SubtleCrypto SHA-384 of 'abc'"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (WebCrypto): SubtleCrypto SHA-512 of 'abc'. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.crypto.digest"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_crypto_sha512_abc",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/crypto?case=sha512_abc"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_crypto_sha512_abc",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "SubtleCrypto SHA-512 of 'abc'"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for @property at-rule registering a typed custom property whose var() resolves to its initial-value when unset. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/css_at_property`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.property.at_property"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "css_at_property",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/css_at_property"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_css_at_property",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "@property at-rule initial-value resolution"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for Cascade origin: a !important class declaration overriding a higher-specificity #id normal declaration. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/css_cascade_important`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade.important"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "css_cascade_important",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/css_cascade_important"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_css_cascade_important",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "cascade !important overrides higher specificity"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for A constructed CSSStyleSheet (replaceSync) applied via a shadow root's adoptedStyleSheets. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/css_constructable`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.constructable_stylesheets"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "css_constructable",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/css_constructable"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_css_constructable",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "constructable stylesheet + shadow adoptedStyleSheets"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for :focus-within matching a container after its descendant input is programmatically focused. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/css_focus_within`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.focus_within"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "css_focus_within",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/css_focus_within"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_css_focus_within",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":focus-within matches ancestor of focused element"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for :has(> :is(.a,.b)) relative child combinator with a nested forgiving selector list. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/css_has_forgiving`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.has_forgiving"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "css_has_forgiving",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/css_has_forgiving"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_css_has_forgiving",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":has() with nested forgiving :is() child combinator"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for :is() taking the specificity of its most specific argument (:is(#id,.class) beats a bare .class rule). The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/css_is_specificity`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.is_specificity"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "css_is_specificity",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/css_is_specificity"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_css_is_specificity",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":is() specificity equals its most specific argument"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for :nth-child(2n of .x) counting only the .x-filtered sibling subset (An+B of S syntax). The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/css_nth_child_of`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.nth_child_of"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "css_nth_child_of",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/css_nth_child_of"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_css_nth_child_of",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":nth-child(An+B of S) filtered structural match"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for Registered custom property inheritance: inherits:true flows to a child while inherits:false falls back to initialValue. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/css_property_inherits`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.property.inherits"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "css_property_inherits",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/css_property_inherits"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_css_property_inherits",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "registered property inherits:true vs false"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for CSS.registerProperty with syntax '<length>' accepting a valid length but falling back to initialValue for an invalid value. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/css_register_syntax`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.property.register_syntax"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "css_register_syntax",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/css_register_syntax"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_css_register_syntax",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CSS.registerProperty <length> typed validation + initial fallback"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for :where() contributing zero specificity so a plain type selector overrides :where(#id,.class). The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/css_where_zero`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.selectors.where"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "css_where_zero",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/css_where_zero"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_css_where_zero",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": ":where() contributes zero specificity"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for AbortController/AbortSignal: a {signal} listener is removed on abort(); signal.aborted and signal.reason reflect the abort. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/dom_abortsignal`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.abort_signal"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dom_abortsignal",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/dom_abortsignal"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_dom_abortsignal",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "AbortSignal removes listener on abort; aborted/reason state"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for Document.adoptNode moving a node's ownerDocument (removing it from the source) vs importNode copying it. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/dom_adoptnode`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.adopt_node"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dom_adoptnode",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/dom_adoptnode"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_dom_adoptnode",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "adoptNode / importNode cross-document ownership"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for Node.cloneNode(true) deep copy vs cloneNode(false) shallow copy, attribute copying, and disconnected result. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/dom_clonenode`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.clone_node"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dom_clonenode",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/dom_clonenode"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_dom_clonenode",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "cloneNode deep vs shallow semantics"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for DOMParser producing a parsererror element for malformed application/xml while text/html auto-corrects without error. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/dom_domparser`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.dom_parser"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dom_domparser",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/dom_domparser"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_dom_domparser",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "DOMParser XML parsererror vs HTML tolerance"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for CustomEvent propagation ordering across capture, at-target, and bubble phases on nested nodes. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/dom_event_order`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.events.propagation_order"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dom_event_order",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/dom_event_order"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_dom_event_order",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "CustomEvent capture/target/bubble listener order"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for Event.stopPropagation preventing ancestor delivery while remaining same-target listeners still fire. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/dom_event_stopprop`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.events.stop_propagation"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dom_event_stopprop",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/dom_event_stopprop"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_dom_event_stopprop",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "stopPropagation blocks ancestors but not same-target listeners"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for MutationObserver attributes with attributeFilter restricting records and attributeOldValue capturing prior values. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/dom_mo_attributes`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.mutation_observer.attributes"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dom_mo_attributes",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/dom_mo_attributes"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_dom_mo_attributes",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "MutationObserver attributeFilter + attributeOldValue"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for MutationObserver characterData with characterDataOldValue recording successive text-node edits. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/dom_mo_chardata`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.mutation_observer.character_data"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dom_mo_chardata",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/dom_mo_chardata"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_dom_mo_chardata",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "MutationObserver characterData oldValue records"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for MutationObserver childList delivering one record per mutation in a single microtask with addedNodes/removedNodes. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/dom_mo_childlist`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.mutation_observer.child_list"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dom_mo_childlist",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/dom_mo_childlist"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_dom_mo_childlist",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "MutationObserver childList added/removed record aggregation"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for NodeIterator over SHOW_TEXT emitting non-whitespace text nodes in document order. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/dom_nodeiterator`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.traversal.node_iterator"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dom_nodeiterator",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/dom_nodeiterator"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_dom_nodeiterator",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "NodeIterator SHOW_TEXT traversal with acceptNode filter"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for Range.extractContents splitting partially selected text nodes across element boundaries and collapsing the range. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/dom_range_extract`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.range.extract_contents"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dom_range_extract",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/dom_range_extract"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_dom_range_extract",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Range.extractContents partial text-node boundary split"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for Range.surroundContents wrapping a fully selected node in a new parent element. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/dom_range_surround`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.range.surround_contents"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dom_range_surround",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/dom_range_surround"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_dom_range_surround",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "Range.surroundContents wrapping semantics"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for HTMLTemplateElement.content as a disconnected DocumentFragment whose contained script does not execute. The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/dom_template_content`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.template_content"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dom_template_content",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/dom_template_content"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_dom_template_content",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "template.content inert DocumentFragment and inert script"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for TreeWalker acceptNode returning FILTER_REJECT (prune subtree) vs FILTER_SKIP (skip node, keep children). The scripted CDP driver observes page output from the self-hosted fixture `/v2/wpt_b/dom_treewalker_filter`, and the server-side grader compares it with the Chrome-baked expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.dom.traversal.tree_walker"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dom_treewalker_filter",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_b/dom_treewalker_filter"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_b",
"version.v0_2"
],
"task_id": "v2_wpt_dom_treewalker_filter",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "TreeWalker acceptNode FILTER_REJECT vs FILTER_SKIP"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (encoding): TextDecoder strips UTF-8 BOM. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.encoding.text_decoder"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_enc_bom_strip",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/enc?case=bom_strip"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_enc_bom_strip",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "TextDecoder strips UTF-8 BOM"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (encoding): TextEncoder.encodeInto read/written counts. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.encoding.encode_into"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_enc_encode_into",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/enc?case=encode_into"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_enc_encode_into",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "TextEncoder.encodeInto read/written counts"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (encoding): TextEncoder.encoding is utf-8. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.encoding.text_encoder"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_enc_encoding_prop",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/enc?case=encoding_prop"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_enc_encoding_prop",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "TextEncoder.encoding is utf-8"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (encoding): TextDecoder fatal mode throws on invalid byte. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.encoding.fatal"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_enc_fatal_throws",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/enc?case=fatal_throws"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_enc_fatal_throws",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "TextDecoder fatal mode throws on invalid byte"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (encoding): TextDecoder label aliases canonicalize. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.encoding.labels"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_enc_labels",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/enc?case=labels"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_enc_labels",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "TextDecoder label aliases canonicalize"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (encoding): TextDecoder non-fatal replacement char. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.encoding.text_decoder"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_enc_nonfatal_replacement",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/enc?case=nonfatal_replacement"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_enc_nonfatal_replacement",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "TextDecoder non-fatal replacement char"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (encoding): TextDecoder stream splits a multibyte char. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.encoding.stream"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_enc_stream_split",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/enc?case=stream_split"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_enc_stream_split",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "TextDecoder stream splits a multibyte char"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (encoding): TextEncoder replaces lone surrogate. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.encoding.text_encoder"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_enc_surrogate_replace",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/enc?case=surrogate_replace"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_enc_surrogate_replace",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "TextEncoder replaces lone surrogate"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (encoding): TextEncoder UTF-8 of astral (4-byte) char. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.encoding.text_encoder"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_enc_utf8_astral",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/enc?case=utf8_astral"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_enc_utf8_astral",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "TextEncoder UTF-8 of astral (4-byte) char"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (storage): localStorage coerces value to string. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.local_storage"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_store_coerce_string",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/store?case=coerce_string"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_store_coerce_string",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "localStorage coerces value to string"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (storage): document.cookie stores multiple cookies. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.cookies"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_store_cookie_multiple",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/store?case=cookie_multiple"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_store_cookie_multiple",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "document.cookie stores multiple cookies"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (storage): document.cookie write then read. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.cookies"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_store_cookie_readwrite",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/store?case=cookie_readwrite"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_store_cookie_readwrite",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "document.cookie write then read"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (storage): storage event fires in another same-origin context. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.storage_event"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_store_event_fires",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/store?case=event_fires"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_store_event_fires",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "storage event fires in another same-origin context"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (storage): localStorage key() semantics. The fixture asserts the actual WPT storage_key contract in-page — every in-range index returns a distinct existing key, out-of-range returns null, and updating an existing key's value preserves the enumeration order — then reports the sorted key set. No specific enumeration order is required.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.local_storage"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_store_key_order",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/store?case=key_order"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_store_key_order",
"task_version": 2,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "localStorage key()/length contract (order-insensitive)"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (storage): localStorage set/get/length/key/removeItem. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.local_storage"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_store_local_setget",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/store?case=local_setget"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_store_local_setget",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "localStorage set/get/length/key/removeItem"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (storage): localStorage overwrite keeps length. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.local_storage"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_store_overwrite_length",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/store?case=overwrite_length"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_store_overwrite_length",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "localStorage overwrite keeps length"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (storage): localStorage removeItem on missing key is a no-op. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.local_storage"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_store_removeitem_missing",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/store?case=removeitem_missing"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_store_removeitem_missing",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "localStorage removeItem on missing key is a no-op"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (storage): sessionStorage is independent of localStorage. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.storage.session_storage"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_store_session_local_independent",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/store?case=session_local_independent"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_store_session_local_independent",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "sessionStorage is independent of localStorage"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (url): URL normalizes backslash in special scheme. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.url.parse"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_url_backslash",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/url?case=backslash"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_url_backslash",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "URL normalizes backslash in special scheme"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (url): URL drops default port 80. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.url.port"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_url_default_port_http",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/url?case=default_port_http"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_url_default_port_http",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "URL drops default port 80"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (url): URL empty query preserves trailing ?. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.url.parse"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_url_empty_query",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/url?case=empty_query"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_url_empty_query",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "URL empty query preserves trailing ?"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (url): URL IDN host converts to punycode. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.url.idn"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_url_idn_punycode",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/url?case=idn_punycode"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_url_idn_punycode",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "URL IDN host converts to punycode"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (url): URL lowercases host and scheme. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.url.parse"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_url_lowercase_host",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/url?case=lowercase_host"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_url_lowercase_host",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "URL lowercases host and scheme"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (url): URL non-special scheme keeps opaque path. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.url.parse"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_url_nonspecial_opaque",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/url?case=nonspecial_opaque"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_url_nonspecial_opaque",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "URL non-special scheme keeps opaque path"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (url): URL percent-encodes space in path. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.url.percent_encoding"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_url_pct_space_path",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/url?case=pct_space_path"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_url_pct_space_path",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "URL percent-encodes space in path"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (url): URL protocol-relative //host resolution. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.url.relative"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_url_protocol_relative",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/url?case=protocol_relative"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_url_protocol_relative",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "URL protocol-relative //host resolution"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (url): URL relative resolution (./ dot segment). The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.url.relative"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_url_rel_dot",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/url?case=rel_dot"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_url_rel_dot",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "URL relative resolution (./ dot segment)"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (url): URL relative resolution (../ dot segments). The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.url.relative"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_url_rel_dotdot",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/url?case=rel_dotdot"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_url_rel_dotdot",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "URL relative resolution (../ dot segments)"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (url): URLSearchParams getAll on repeated key. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.url.search_params"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_url_sp_getall",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/url?case=sp_getall"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_url_sp_getall",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "URLSearchParams getAll on repeated key"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (url): URLSearchParams decodes + as space. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.url.search_params"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_url_sp_plus_space",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/url?case=sp_plus_space"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_url_sp_plus_space",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "URLSearchParams decodes + as space"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (url): URLSearchParams set replaces all values. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.url.search_params"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_url_sp_set_replace",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/url?case=sp_set_replace"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_url_sp_set_replace",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "URLSearchParams set replaces all values"
},
{
"artifact_profile": "l2_standard",
"description": "WPT-derived L2 mechanism probe (url): URLSearchParams sort orders keys. The self-hosted fixture computes the WHATWG/Web-platform result in-page and writes it to `#out`; the scripted CDP driver reads it and the server-side grader compares against the Chrome-oracle answer. Rewritten from WPT testharness assertions (not run verbatim).",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 5000,
"wait_ms": 250
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.url.search_params"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "v2_wpt_url_sp_sort",
"kind": "self_hosted_fixture",
"url": "/v2/wpt_a/url?case=sp_sort"
},
"subset_id": "l2.web_platform",
"tags": [
"family.web_platform_a",
"version.v0_2"
],
"task_id": "v2_wpt_url_sp_sort",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "URLSearchParams sort orders keys"
}
]
}
{
"cases": [
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web-platform semantics probe for ::before generated content. The driver reads output from the self-hosted fixture `/fixtures/l2_round3/dA_e_before_content.html`, and the server-side grader compares it with the checked-in expected answer. Current probe details: output prefixed 'LABEL:'; expected token NEW_ARRIVAL inside ::before content string.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dA_e_before_content",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dA_e_before_content.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dA_e_before_content",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "::before generated content"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for getComputedStyle:color cascade. The scripted CDP driver observes page output from the self-hosted fixture `/fixtures/l2_round3/dA_e_color_gate.html`, and the server-side grader compares it with the expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {
"PROBE_SPEC": {
"read": "textContent",
"selector": "#out",
"settle": "nonempty",
"settle_timeout_ms": 3000,
"wait_ms": 200
}
},
"kind": "node_cdp_probe",
"script": "runner/scripts/l2_fixture_probe.js"
},
"features": [
"web.css.cascade",
"web.css.computed_style"
],
"grader": {
"checks": [
"answer_matches_expected"
],
"endpoint": "/__grade__/expected_answer",
"kind": "server_side"
},
"layer": "L2",
"scene": {
"fixture_id": "dA_e_color_gate",
"kind": "self_hosted_fixture",
"url": "/fixtures/l2_round3/dA_e_color_gate.html"
},
"subset_id": "l2.web_platform",
"tags": [
"version.v0_1"
],
"task_id": "r3_dA_e_color_gate",
"task_version": 1,
"timeouts": {
"hard_kill_ms": 30000,
"task_ms": 20000
},
"title": "getComputedStyle:color cascade"
},
{
"artifact_profile": "l2_standard",
"description": "Runs a fixture-backed Web platform semantics probe for getComputedStyle:font-size cascade. The scripted CDP driver observes page output from the self-hosted fixture `/fixtures/l2_round3/dA_e_fontsize_gate.html`, and the server-side grader compares it with the expected answer.",
"driver": {
"connect_env": "BROWSER_WS",
"env": {