TermiGen: stages 1, 3, 5 and teacher results; 3,369 retained tasks

**No gold/reference-solution checks were run: TermiGen provides no gold solutions. Broken tasks will likely remain in this dataset and will be filtered out in the next version.** Teacher passes are evidence of verifier acceptance, not proof that a task or its tests are correct.

## TermiGen

Keeps **3,369 tasks** and archives **197** from [ucsb-mlsec/terminal-bench-env](https://github.com/ucsb-mlsec/terminal-bench-env/tree/03bddac74ada2fa344df0e93b4a3ace2034294d1), covering all **3,566 source tasks**. This PR includes the task payloads, archived tasks with their reasons, and the supporting validation and teacher reports.

The earlier [stage-3 publication (PR #8)](https://huggingface.co/datasets/FWeindel/validated-tasks/discussions/8) retained 3,506 tasks and archived 60. This PR adds static and no-op results, excluding a further **137 tasks**: 133 with absolute-path findings, 2 with CPU-count findings, and 2 whose verifier passed with no answer. No previously archived task is restored.

Filtering starts with all 3,566 source tasks and applies stages 1, 3 and 5 in order. Each excluded task is counted once, at its first applicable exclusion. The counts use the full validation run and recorded environment reruns.

| Check | Tasks entering this step | Newly archived | Tasks retained after this step |
| --- | ---: | ---: | ---: |
| 1: Static checks, with the recorded advisory exceptions | 3,566 | 141 | 3,425 |
| 3: Environment build, startup and inspected files | 3,425 | 54 | 3,371 |
| 5: No-op verification | 3,371 | 2 | 3,369 |

**Only 3,336 retained tasks passed all three measured validation phases.** The remaining 33 are retained as unresolved under the publisher's existing policy: 28 verifiers executed no tests and 5 verifiers timed out. These are not passing no-op results. Stage 4 was not run because there are no reference solutions.

Static results use the recorded training-profile exclusions; pip-pinning and test-file-reference findings remain advisory. The detailed policy and per-task findings are preserved in the [run record](runs/termigen-phases-1-3-5-6-7-preview-20261014.json).

## Archived tasks

The first applicable exclusion accounts for each task once: **141 at stage 1, 54 at stage 3, and 2 at stage 5**. Stage 3 still has the same 60 archived outcomes as PR #8; six of those tasks are counted at stage 1 in this combined report.

| Archive label | Tasks | Explanation |
| --- | ---: | --- |
| build-command-unavailable | 1 | The build could not find a required shell command. |
| container-start-failed | 7 | Apptainer could not start the environment; the saved logs include fakeroot startup diagnostics. |
| copy-destination-collision | 2 | Build-context copying collided with a file already present at the destination. |
| dependency-build-failed | 12 | Environment setup or dependency installation failed. Per-task errors are preserved in the linked stage report. |
| environment-check-incomplete | 9 | Environment startup or inspection did not finish successfully; the report contains no error text. These tasks remain archived under the recorded phase 3 policy. |
| image-preparation-failed | 3 | CUDA, Lean and Ray image builds failed with dpkg package-configuration errors in all three reruns. Later attempts encountered missing images and did not rebuild them. |
| missing-copy-source | 6 | The Dockerfile references a build-context file that is absent from the task payload. |
| no-op-reward-one | 2 | no-answer reward [1.0] instead of 0 |
| runtime-file-tree-mismatch | 12 | Expected task files were missing or differed from their recorded contents inside the running environment. |
| static:compose-host-binds | 3 | static checks failed: check-compose-host-binds.sh |
| static:nproc | 3 | static checks failed: check-nproc.sh |
| static:task-absolute-path | 135 | static checks failed: check-task-absolute-path.sh |
| unsupported-compose-sidecars | 2 | The Apptainer backend cannot run the required Docker Compose sidecars; all three reruns remained blocked. |

The eight previously unresolved environment tasks remain archived after their recorded reruns: five require unsupported Compose sidecars and three failed image preparation. Later attempts on those three encountered missing images rather than rebuilding them. Nine other environment checks have incomplete startup or inspection evidence and remain archived under the reviewed stage-3 policy. These outcomes do not establish whether a valid solution exists.

## Repairs

Our patcher modified **2,219 of the original 3,566 tasks**. A task can have multiple changes, so the counts below overlap.

| Change | Tasks affected | Explanation |
| --- | ---: | --- |
| deferred-helper | 4 | Reconstruct supplied build helpers when deferred installation runs before COPY payloads are available. |
| env-expansion | 51 | Resolve supported Docker ENV substitutions so inherited command paths remain usable. |
| initial-file | 3 | Materialize initial files from supplied payloads so they remain visible after container mounts. |
| tmp-fixture | 158 | Relocate supplied fixtures hidden by temporary-directory mounts and update their path references. |
| wildcard-copy | 1 | Expand build-context wildcard copies into explicit file copies. |
| workspace-copy | 2,058 | Repair COPY destinations and preserve renamed files through byte-identical aliases. |

The patcher retained all source tasks, including 11 for which no safe automatic patch was available. Their patch predictions and validation outcomes remain recorded separately.

## Teacher results: pass@k

Both models used **Terminus 2** on the same **3,336 tasks**, with **four attempts per task and model** (13,344 attempts each; 26,688 total). Stages 6 and 7 are informational and do not exclude tasks. The 33 tasks with unresolved no-op results were not included in teacher generation.

| Model | Pass@1 | Pass@2 | Pass@3 | Pass@4 | Tasks solved at least once |
| --- | ---: | ---: | ---: | ---: | ---: |
| Qwen3-Coder-30B-A3B-Instruct | 48.46% | 57.42% | 61.38% | 63.64% | 2,123 / 3,336 |
| Qwen3-30B-A3B-Instruct-2507 | 37.54% | 47.41% | 52.33% | 55.46% | 1,850 / 3,336 |

Pass@k estimates success within k attempts from the four recorded attempts per task. Missing rewards count as unsolved in these rates; the underlying missing rewards remain missing. At k=4, the rate is the fraction of tasks solved at least once. The models solved **2,277 distinct tasks** between them, including 154 solved only by the Instruct model; that combined coverage is not a single-model pass@k result.

| Model | Agent timeouts | Missing rewards | Missing traces |
| --- | ---: | ---: | ---: |
| Qwen3-Coder-30B-A3B-Instruct | 414 | 142 | 32 |
| Qwen3-30B-A3B-Instruct-2507 | 1,816 | 179 | 32 |

All planned result records are present, but some rewards and traces are missing. Timeout and execution-error findings are included in the results. The Instruct run completed all 17 stage-6/stage-7 batches within their 4-hour limits. Its Slurm failure statuses reflect recorded trial findings, not interrupted dataset coverage.

## Runtime and images

Environment validation used Apptainer on Helma. The original validation evidence records Python 3.12.7 and Harbor 0.8.1; the later eight-task environment reruns used Python 3.12.14, Apptainer 1.5.4, 48 CPUs, 64 GiB per job and concurrency 2. These are historical runs, not a fresh full-dataset reproduction.

The Instruct teacher run used one H200 and 32 CPUs per job, concurrency 16, a 65,536-token context and a 16,384-token reply limit. Jobs took 2h47m to 3h29m and consumed 53.57 GPU-hours in total. The Coder run used concurrency 10 and consumed 70.24 GPU-hours; model and concurrency both changed, so this is not an isolated concurrency-speedup comparison.

Existing cached images were reused for teacher generation. **No cached image bundles are included in this PR** because the legacy cache lacks the release provenance required by the current publisher. Task payloads and Dockerfiles are included.

## Files and evidence

The files below are included in this PR; the patch-script link points to the recorded GitHub commit. The teacher reports contain summaries and per-attempt metrics; the raw conversations and trial logs remain in the original run archives referenced by the provenance files. The image manifest records missing release bundles; it does not contain runnable images.

- [Retained tasks](termigen/tasks.parquet) and [archived tasks with reasons](termigen/archive.parquet).
- [Run record and per-task findings](runs/termigen-phases-1-3-5-6-7-preview-20261014.json).
- Validation reports: [stage 1](evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-1.json), [reviewed stage 3](evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-3.json), and [stage 5](evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-5.json).
- [Teacher comparison and pass@1 through pass@4](evidence/termigen-phases-1-3-5-6-7-preview-20261014/teacher-comparison.json).
- Instruct results: [stage 6](evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-6.json), [stage 7](evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-7.json), and [run summary](evidence/termigen-phases-1-3-5-6-7-preview-20261014/final-outcome-summary.json).
- [Coder outcome and trace summary](evidence/termigen-phases-1-3-5-6-7-preview-20261014/coder-outcome-summary.json).
- [Input evidence and teacher-batch provenance](evidence/termigen-phases-1-3-5-6-7-preview-20261014/input-provenance.json).
- [Patch comparison with upstream](evidence/termigen-phases-1-3-5-6-7-preview-20261014/patch-manifest.json), [patcher reproduction check](evidence/termigen-phases-1-3-5-6-7-preview-20261014/patcher-reproduction.json), and [patch script at the recorded commit](https://github.com/MLI-lab/OT-next-data/blob/f06623d236ee24515ef479446e029f79463559e3/data/termigen/patch.py).
- [Image release manifest](images/manifests/termigen-phases-1-3-5-6-7-preview-20261014.json).

.gitattributes

@@ -58,3 +58,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
+evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-1.json filter=lfs diff=lfs merge=lfs -text
+evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-3.json filter=lfs diff=lfs merge=lfs -text
+evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-7.json filter=lfs diff=lfs merge=lfs -text

evidence/termigen-phases-1-3-5-6-7-preview-20261014/aggregate-contract.json

@@ -0,0 +1,889 @@
+{
+  "schema_version": 2,
+  "created_at": "2026-10-07T17:39:01.945729+00:00",
+  "dataset": {
+    "source": "/hnvme/workspace/y500bb10-termigen-audit/phase1-static-20261007-152318/full/contract.stage1-tasks",
+    "revision": "sha256:9fdcf3605517fa4e547669d8256f06d528d9bb05e6c6ea3a5a515f7916cf06b9",
+    "source_files": [
+      {
+        "path": "/hnvme/workspace/y500bb10-termigen-audit/phase1-static-20261007-152318/full/contract.stage1-tasks/0000.parquet",
+        "sha256": "7ce3285f0c001ab0ca4b687d9611797789012c6c9ce72046876eb9a72d8a72fb"
+      }
+    ],
+    "preselection": null,
+    "normalization": {
+      "operation": "canonical-timeout-anticheat-suffix-v1",
+      "original_path": "/hnvme/workspace/y500bb10-termigen-audit/phase0-final-v3/tasks.parquet",
+      "description": "Our extension to the upstream Terminal-Bench suffix check appends the sentence when missing and overwrites the saved instruction.md in the validation copy; an existing canonical suffix is updated without duplication. Original inputs are unchanged.",
+      "sentence_template": "You have N seconds to complete this task. Do not cheat by using online solutions or hints specific to this task.",
+      "timeout_source": "[agent].timeout_sec",
+      "saved_dataset": "/hnvme/workspace/y500bb10-termigen-audit/phase1-static-20261007-152318/full/contract.stage1-tasks",
+      "original_manifest_sha256": "489f57f05a6705572fa7dbcaf37fdfc29c4ac4a2b3d27b47c7d88f8ee7e5153a",
+      "original_source_files": [
+        {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/phase0-final-v3/tasks.parquet",
+          "sha256": "5c66bf3eac87d0948166c4fc45400aff4cac9173aed6cb5e3788ad3dc1ce30cd"
+        }
+      ],
+      "original_selection": {
+        "limit": null,
+        "task_id_range": null
+      }
+    },
+    "selection": {
+      "method": "all",
+      "limit": null,
+      "task_id_range_inclusive": null,
+      "order": "lexicographic IDs for ranges; otherwise sorted Parquet paths then row order, or sorted task directories",
+      "implementation": "validation/contract.py:inventory (hash recorded in implementation)"
+    }
+  },
+  "stages": [
+    1,
+    3,
+    5
+  ],
+  "success_criteria": {
+    "minimum_tasks": 3566,
+    "oracle_reward": 1,
+    "nop_reward": 0,
+    "reward_key": "reward",
+    "attempts": 1,
+    "static_checks": [
+      "check-compose-host-binds.sh",
+      "check-dockerfile-platform.sh",
+      "check-dockerfile-references.sh",
+      "check-dockerfile-sanity.sh",
+      "check-instruction-suffix.sh",
+      "check-no-cheat-dir.sh",
+      "check-nproc.sh",
+      "check-pip-pinning.sh",
+      "check-resource-sizes.sh",
+      "check-task-absolute-path.sh",
+      "check-task-timeout.sh",
+      "check-test-file-references.sh",
+      "check-trial-network-fetch.sh",
+      "check-verifier-tooling-baked.sh",
+      "check_ai_detection.py"
+    ],
+    "static_exclusions": {
+      "check-task-changelog.sh": "training tasks do not require Terminal-Bench PR changelogs",
+      "check-test-sh-sanity.sh": "training permits shared system Python and image-installed verifier dependencies; uv is not required, and stages 4/5 check execution",
+      "check-separate-verifier.sh": "training permits grading in the agent environment, including installed packages and changed system state; separate verification is optional",
+      "check-allow-internet.sh": "network policy is declared by the task and checked at runtime",
+      "check-no-allow-internet-true.sh": "explicit task network policies are allowed",
+      "check-pytest-version.sh": "training tasks may use other pinned pytest/CTRF versions",
+      "check-gpu-types.sh": "Modal GPU naming conventions do not apply to the Helma execution profile",
+      "check-canary.sh": "training tasks do not require a benchmark canary string",
+      "check-instruction-headings.sh": "Markdown headings are allowed in training instructions",
+      "check-task-fields.sh": "training tasks do not require TB author GitHub metadata, taxonomy or explanation sections",
+      "check-task-package-name.sh": "training task package names need not use the terminal-bench namespace",
+      "check-task-slug.sh": "training task names may contain more than three hyphen-separated tokens",
+      "rubric_review.py": "LLM rubric review is opt-in via stage 2, never a default static check"
+    },
+    "conditional_checks": {
+      "check_ai_detection.py": "only with GPTZERO_API_KEY; missing key or upstream API-unavailable skip is non-failing"
+    },
+    "failure_policy": "collect all findings; fail acceptance on any selected failure/error/skip or insufficient coverage"
+  },
+  "execution_profile": {
+    "agent_settings": {
+      "settings": {
+        "model_name": null,
+        "max_turns": null,
+        "parser_name": "json",
+        "api_base": null,
+        "temperature": null,
+        "reasoning_effort": null,
+        "collect_rollout_details": false,
+        "session_id": null,
+        "enable_summarize": true,
+        "proactive_summarization_threshold": 8000,
+        "max_thinking_tokens": null,
+        "model_info": null,
+        "trajectory_config": null,
+        "trajectory_dump_cadence": null,
+        "tmux_pane_width": 160,
+        "tmux_pane_height": 40,
+        "store_all_messages": false,
+        "enable_episode_logging": false,
+        "interleaved_thinking": false,
+        "suppress_max_turns_warning": false,
+        "use_responses_api": false,
+        "llm_backend": "litellm",
+        "llm_kwargs": null,
+        "llm_call_kwargs": null,
+        "extra_body": null,
+        "extra_env": null,
+        "turn_callback": null,
+        "batch_turn_execs": true,
+        "batch_exec_timeout_margin_sec": 180.0
+      },
+      "defaults_resolved": true,
+      "source_sha256": {
+        "terminus_2.py": "256df8b60461c9326468f843e5c3a0f483d03b473a47d6b5a1d442ab0b6d48cd",
+        "terminus_json_plain_parser.py": "b20ad9b8931d4f92c31aa588126a342d9cc6a59396f3ef21243bd077895d9022",
+        "terminus_xml_plain_parser.py": "11ff85d848d4ada5874644b5113b08dbcc53de81deb6ba4e014ca6fce92435fa",
+        "templates/terminus-json-plain.txt": "89a3dc3a15752b748a99fb907c9251ab31a464e13386583fb48476b078b54a34",
+        "templates/terminus-xml-plain.txt": "721c0ada54921cbb7bb9f817f2306c79ff4ba3381d08be7eb97fbeb0d99d298d",
+        "templates/timeout.txt": "32bf9aa7b157a6a0e0ba0d33b81e59dd67204276f158afc1fe6cb684ced68872"
+      }
+    },
+    "backend": "apptainer",
+    "architecture": "x86_64",
+    "partition_policy": "auto",
+    "allocation_policy": "auto: CPU for external models; h200 for local models or explicit GPUs; h200 fallback if CPU partition has no usable nodes",
+    "scheduler_target": "never",
+    "submission_host": "h35-19",
+    "network": "host",
+    "network_guarantee": "host networking through cluster HTTP proxy",
+    "privileges": "fakeroot requested, host UID unprivileged; bridge may fall back without fakeroot",
+    "mount_policy": "per-trial workdir, tests, logs, home and tmp; site hostfs/bind-paths/cwd disabled; baked verifier tests preserved; explicit read-only host DNS bind only in host network mode",
+    "profile_scope": "declared bridge policy including stated fallbacks; does not certify offline or rootless behavior",
+    "dependency_policy": "execute shipped Dockerfiles; mutable tags and unpinned installs are not rewritten; reuse content-keyed image cache unless force_build; installed agents may download dependencies",
+    "environment_kwargs": {},
+    "force_build": false,
+    "trial_cpus": null,
+    "trial_memory_mb": null,
+    "agent": "terminus-2",
+    "model": null,
+    "review_agent": null,
+    "review_model": null,
+    "analysis_agent": null,
+    "analysis_model": null,
+    "review_defaults_source": null,
+    "agent_kwargs": {},
+    "proposal_review": false,
+    "proposal_model": null,
+    "review_local": false,
+    "serve_weights": null,
+    "serve_image": null,
+    "serve_context": 32768,
+    "local_model_assets": null,
+    "serve_model": null,
+    "api_base": null,
+    "validation_container_reuse": "same-task stages 3,5,4; logs cleared, filesystem state retained",
+    "container_start_concurrency": 2,
+    "container_start_interval": 0.25,
+    "static_concurrency": 2,
+    "static_concurrency_policy": "cap at allocated CPU budget; no separate verifier reservation",
+    "concurrency": 2,
+    "concurrency_policy": "cap at allocation CPU budget including separate verifiers; record effective value in resources.json",
+    "cpus": 48,
+    "gpus": null,
+    "serve_cpus": 4,
+    "serve_memory_mb": 32768,
+    "max_num_seqs": null
+  },
+  "runtime_dependencies": {
+    "python": "3.12.7",
+    "packages": [
+      [
+        "Deprecated",
+        "1.3.1"
+      ],
+      [
+        "Jinja2",
+        "3.1.6"
+      ],
+      [
+        "MarkupSafe",
+        "3.0.3"
+      ],
+      [
+        "PyJWT",
+        "2.15.0"
+      ],
+      [
+        "PyYAML",
+        "6.0.3"
+      ],
+      [
+        "Pygments",
+        "2.21.0"
+      ],
+      [
+        "StrEnum",
+        "0.4.15"
+      ],
+      [
+        "aiofiles",
+        "25.1.0"
+      ],
+      [
+        "aiohappyeyeballs",
+        "2.7.1"
+      ],
+      [
+        "aiohttp",
+        "3.14.3"
+      ],
+      [
+        "aiohttp-retry",
+        "2.9.1"
+      ],
+      [
+        "aiosignal",
+        "1.4.0"
+      ],
+      [
+        "annotated-doc",
+        "0.0.5"
+      ],
+      [
+        "annotated-types",
+        "0.8.0"
+      ],
+      [
+        "anyio",
+        "4.15.1"
+      ],
+      [
+        "attrs",
+        "26.1.0"
+      ],
+      [
+        "bidict",
+        "0.24.1"
+      ],
+      [
+        "boto3",
+        "1.43.101"
+      ],
+      [
+        "botocore",
+        "1.43.101"
+      ],
+      [
+        "certifi",
+        "2026.7.22"
+      ],
+      [
+        "cffi",
+        "2.1.1"
+      ],
+      [
+        "charset-normalizer",
+        "3.5.1"
+      ],
+      [
+        "click",
+        "8.5.0"
+      ],
+      [
+        "contourpy",
+        "1.4.0"
+      ],
+      [
+        "cryptography",
+        "50.0.1"
+      ],
+      [
+        "cycler",
+        "0.12.1"
+      ],
+      [
+        "daytona",
+        "0.203.0"
+      ],
+      [
+        "daytona_analytics_api_client",
+        "0.203.0"
+      ],
+      [
+        "daytona_analytics_api_client_async",
+        "0.203.0"
+      ],
+      [
+        "daytona_api_client",
+        "0.203.0"
+      ],
+      [
+        "daytona_api_client_async",
+        "0.203.0"
+      ],
+      [
+        "daytona_toolbox_api_client",
+        "0.203.0"
+      ],
+      [
+        "daytona_toolbox_api_client_async",
+        "0.203.0"
+      ],
+      [
+        "deprecation",
+        "2.1.0"
+      ],
+      [
+        "dirhash",
+        "0.5.0"
+      ],
+      [
+        "distro",
+        "1.9.0"
+      ],
+      [
+        "fastapi",
+        "0.141.1"
+      ],
+      [
+        "fastuuid",
+        "0.14.0"
+      ],
+      [
+        "filelock",
+        "4.0.3"
+      ],
+      [
+        "fonttools",
+        "4.66.0"
+      ],
+      [
+        "frozenlist",
+        "1.8.0"
+      ],
+      [
+        "fsspec",
+        "2026.9.0"
+      ],
+      [
+        "googleapis-common-protos",
+        "1.75.3"
+      ],
+      [
+        "h11",
+        "0.16.0"
+      ],
+      [
+        "h2",
+        "4.4.1"
+      ],
+      [
+        "harbor",
+        "0.8.1"
+      ],
+      [
+        "hf-xet",
+        "1.6.0"
+      ],
+      [
+        "hpack",
+        "4.2.0"
+      ],
+      [
+        "httpcore",
+        "1.0.9"
+      ],
+      [
+        "httpcore2",
+        "2.13.1"
+      ],
+      [
+        "httpx",
+        "0.28.1"
+      ],
+      [
+        "httpx-ws",
+        "0.9.0"
+      ],
+      [
+        "httpx2",
+        "2.13.1"
+      ],
+      [
+        "huggingface_hub",
+        "1.33.0"
+      ],
+      [
+        "hyperframe",
+        "6.1.0"
+      ],
+      [
+        "idna",
+        "3.20"
+      ],
+      [
+        "importlib_metadata",
+        "8.9.0"
+      ],
+      [
+        "iniconfig",
+        "2.3.0"
+      ],
+      [
+        "jiter",
+        "0.17.0"
+      ],
+      [
+        "jmespath",
+        "1.1.0"
+      ],
+      [
+        "jsonschema",
+        "4.26.0"
+      ],
+      [
+        "jsonschema-specifications",
+        "2025.9.1"
+      ],
+      [
+        "kiwisolver",
+        "1.5.1"
+      ],
+      [
+        "litellm",
+        "1.102.1"
+      ],
+      [
+        "markdown-it-py",
+        "4.2.0"
+      ],
+      [
+        "matplotlib",
+        "3.11.2"
+      ],
+      [
+        "mdurl",
+        "0.1.2"
+      ],
+      [
+        "multidict",
+        "6.9.1"
+      ],
+      [
+        "numpy",
+        "2.5.3"
+      ],
+      [
+        "obstore",
+        "0.11.1"
+      ],
+      [
+        "openai",
+        "2.54.0"
+      ],
+      [
+        "opentelemetry-api",
+        "1.44.0"
+      ],
+      [
+        "opentelemetry-exporter-otlp-proto-common",
+        "1.44.0"
+      ],
+      [
+        "opentelemetry-exporter-otlp-proto-http",
+        "1.44.0"
+      ],
+      [
+        "opentelemetry-instrumentation",
+        "0.65b0"
+      ],
+      [
+        "opentelemetry-instrumentation-aiohttp-client",
+        "0.65b0"
+      ],
+      [
+        "opentelemetry-proto",
+        "1.44.0"
+      ],
+      [
+        "opentelemetry-sdk",
+        "1.44.0"
+      ],
+      [
+        "opentelemetry-semantic-conventions",
+        "0.65b0"
+      ],
+      [
+        "opentelemetry-util-http",
+        "0.65b0"
+      ],
+      [
+        "packaging",
+        "26.3"
+      ],
+      [
+        "pathlib_abc",
+        "0.5.2"
+      ],
+      [
+        "pathspec",
+        "1.1.1"
+      ],
+      [
+        "pillow",
+        "12.3.0"
+      ],
+      [
+        "pip",
+        "26.2.1"
+      ],
+      [
+        "pluggy",
+        "1.6.0"
+      ],
+      [
+        "postgrest",
+        "2.31.0"
+      ],
+      [
+        "propcache",
+        "0.5.4"
+      ],
+      [
+        "protobuf",
+        "7.36.2"
+      ],
+      [
+        "pyarrow",
+        "25.0.1"
+      ],
+      [
+        "pycparser",
+        "3.0"
+      ],
+      [
+        "pydantic",
+        "2.13.5"
+      ],
+      [
+        "pydantic-settings",
+        "2.15.0"
+      ],
+      [
+        "pydantic_core",
+        "2.46.5"
+      ],
+      [
+        "pyparsing",
+        "3.3.3"
+      ],
+      [
+        "pytest",
+        "9.1.1"
+      ],
+      [
+        "python-dateutil",
+        "2.9.0.post0"
+      ],
+      [
+        "python-dotenv",
+        "1.2.3"
+      ],
+      [
+        "python-engineio",
+        "4.14.0"
+      ],
+      [
+        "python-multipart",
+        "0.0.32"
+      ],
+      [
+        "python-socketio",
+        "5.17.0"
+      ],
+      [
+        "realtime",
+        "2.31.0"
+      ],
+      [
+        "referencing",
+        "0.37.0"
+      ],
+      [
+        "regex",
+        "2026.9.10"
+      ],
+      [
+        "requests",
+        "2.34.2"
+      ],
+      [
+        "rich",
+        "15.0.0"
+      ],
+      [
+        "rpds-py",
+        "2026.6.3"
+      ],
+      [
+        "s3transfer",
+        "0.19.2"
+      ],
+      [
+        "scantree",
+        "0.0.4"
+      ],
+      [
+        "shellingham",
+        "1.5.4"
+      ],
+      [
+        "shortuuid",
+        "1.0.13"
+      ],
+      [
+        "simple-websocket",
+        "1.1.0"
+      ],
+      [
+        "six",
+        "1.17.0"
+      ],
+      [
+        "sniffio",
+        "1.3.1"
+      ],
+      [
+        "starlette",
+        "1.7.0"
+      ],
+      [
+        "storage3",
+        "2.31.0"
+      ],
+      [
+        "supabase",
+        "2.31.0"
+      ],
+      [
+        "supabase-auth",
+        "2.31.0"
+      ],
+      [
+        "supabase-functions",
+        "2.31.0"
+      ],
+      [
+        "tenacity",
+        "9.1.4"
+      ],
+      [
+        "tiktoken",
+        "0.14.0"
+      ],
+      [
+        "tokenizers",
+        "0.23.2"
+      ],
+      [
+        "toml",
+        "0.10.2"
+      ],
+      [
+        "tqdm",
+        "4.70.1"
+      ],
+      [
+        "truststore",
+        "0.10.4"
+      ],
+      [
+        "typer",
+        "0.27.2"
+      ],
+      [
+        "typing-inspection",
+        "0.4.4"
+      ],
+      [
+        "typing_extensions",
+        "4.16.0"
+      ],
+      [
+        "universal_pathlib",
+        "0.3.10"
+      ],
+      [
+        "urllib3",
+        "2.8.0"
+      ],
+      [
+        "uvicorn",
+        "0.53.0"
+      ],
+      [
+        "websocket-client",
+        "1.9.2"
+      ],
+      [
+        "websockets",
+        "15.0.1"
+      ],
+      [
+        "wrapt",
+        "2.4.1"
+      ],
+      [
+        "wsproto",
+        "1.3.2"
+      ],
+      [
+        "yarl",
+        "1.25.1"
+      ],
+      [
+        "zipp",
+        "4.1.0"
+      ]
+    ]
+  },
+  "trajectory_input_sha256": null,
+  "upstream": {
+    "terminal-bench": {
+      "url": "https://github.com/harbor-framework/terminal-bench.git",
+      "commit": "1dcda8716784493721921c23e4bc7f7d988b4494",
+      "sparse": [
+        "scripts",
+        "docs",
+        ".github"
+      ]
+    },
+    "harbor-validation": {
+      "url": "https://github.com/marin-community/harbor.git",
+      "commit": "7faf878c14b6d72737579ec936ebf5f74ba3194d",
+      "sparse": [
+        "src",
+        "packages"
+      ]
+    },
+    "harden-v0": {
+      "url": "https://github.com/few-sh/harden-v0.git",
+      "commit": "342b8474e0c0cf96e4a8313fd2e26c7a11d51193"
+    }
+  },
+  "implementation": {
+    "config/clusters.py": "4318d67976d202fbe0c87fd4ffce4d91ea4dd1a11bbd015c3afefc55837f2bac",
+    "config/models.py": "d55d6df2a0aa537e78b6792805dc12878e763115ebc19e195f11f27dc412570b",
+    "data/annotate_dataset/capability_taxonomy.txt": "10d929a31dde33f96858fbd844afc04bc73a40b858dcf5602f48e868d7cd0351",
+    "data/annotate_dataset/domain_taxonomy.txt": "9a5e80ce73531517395b3e51208870a06487fcc391c3eff70d4bcf789ee333cd",
+    "data/annotate_dataset/evaluation_suite.txt": "a6aa12c845a9827268059311fa841a740950c10b3aa341b22cd6f331d527d956",
+    "data/annotate_dataset/prompt_template.txt": "1db5eddfe6b07e3bf9382f63cd96f8c3bd35ffb30446737589b72592c835884e",
+    "harbor_patches/bridge_worker.py": "625e6fa16a82e840e29cdcb62a74ef99e390c03eb62b58fec5122c5f7abf9b1f",
+    "hpc/helma/proxy.sh": "ab27bb020a914f87e7ceaa592c00d505828e9808a5dbf33231421bbf392e9c5b",
+    "hpc/helma/validation.sbatch": "34d0d451663c1b6681f9e17ca4f91ddcf06dfde7b59a7b6d9cefb694535935b0",
+    "hpc/helma/validation_submit.py": "77fdded64d199e9608004816ededcdd5ee10823b0240a8575c341e53bc74e6df",
+    "hpc/helma/validation_worker.py": "fc859467373886863a5d43bcbd784add3a9e042b4afab6cca4823fe3966c4714",
+    "validation/__init__.py": "c2d3352077872dbacd12b4a3746f052fa69d2a7b17235762f5390f1fa9e14488",
+    "validation/annotation.py": "a2a0401afdd46d63bcad8e48841556c5506bad980f8c2391ba59ed174f77b940",
+    "validation/annotation_runtime.py": "fbe13e5454edf2d71fb23c1a4acc105065d7a681c12d6ab9ecc8fad99335c3ec",
+    "validation/checks/__init__.py": "a8699eb188cb9855f0601280b451f2fa4f5e58030aac77f0d56a0d75f1a74ef7",
+    "validation/checks/agent_run.py": "6e4bb8476712abbba2f6f171a480a3964b47f8ae79764fa0bf31d0da7627f01e",
+    "validation/checks/command_metrics.py": "e344064bc3881c0838a9d07425a5588ac5edc12ce2ddbc3245763b4160643a86",
+    "validation/checks/environment.py": "f5e38d74a359aaabbe90ead56132c3998a8bc1b67c8efff76184ac64de5f1daf",
+    "validation/checks/instruction_suffix.py": "cf6c1cf3366be3fb6a5bb9b9a394e9d5b54b5ec077c3b854e2bbc40f23c1707f",
+    "validation/checks/network.py": "57ee8f6cce6fdc702014ca4adc0a8e3c308b3c8ae7012ae06c2a14b3b5da3c4e",
+    "validation/checks/proposal_review.py": "4d0ee52c2ad25276fd9bc214323ddf1fa8bd3f99176e2925bef05e5632a48adc",
+    "validation/checks/review_results.py": "c22457df68dcdfa6ac24e56055612bb3886a6a251feb0393d934b3c6ddbb814e",
+    "validation/checks/reward_metrics.py": "12bcbf2e545e4150a65ddbd2705dc2c49268a5aa3a307e59ad7dec8fd5b96863",
+    "validation/checks/shell_commands.py": "899cf36570d0dbe6b28985956f7cb94bde2b83fb187728bfeb700bd00f6aba37",
+    "validation/checks/trace_metrics.py": "c0011380afa8b4f9ed091367905914d291e769dba1d3cedbd2752fe5f5d9c45c",
+    "validation/contract.py": "c7b7b190f5e3be628b95289980d87ef7733adb03dc50d294efbd12513131ad64",
+    "validation/create_pilot.py": "8d30f45e37d24d64e0726233ee10c1ddc91528fa1fc42f698769a119b81962fd",
+    "validation/data/__init__.py": "aea9b5b7003cefb8631b11ac533ee2c3de8fe332ae56abbbe391bf573461ece4",
+    "validation/data/materialize.py": "9dd88ecaede899a23505a6e5a4c35e4ea9487587f422f07de48095a71f92d0a5",
+    "validation/data/select_smoke_sample.py": "739c48d00b026411c60da1319d6a3a2397e1994c2fcb69f2b806eebd6eb7bd3d",
+    "validation/data/selection.py": "66d77c1d88c7a0ecb2d1233624c98f60b15e7d443f027ef9fa102bfa9558f94b",
+    "validation/dataset_checks.py": "5fad0de982aa6eff54315f2d330ab7c269716161ebad286349965d66eaaaab0c",
+    "validation/publish.py": "352650bbab7e1bf567549f78d6f78ce6312ef768e867dcf1adfc87b8cafe80ce",
+    "validation/report.py": "ac0b709513362f9864720f6ee85b1d072aff399d3a26c7ce8d809b0d974ebfe4",
+    "validation/run.py": "b8852d1701c43984a1fc440fe4e545bc1d511d19b8224edbf5b97f14aae69f1f",
+    "validation/stages/__init__.py": "69006b2aff4beba17bc61e0bea8370f915fbaa5b3620ca91f267de43272c4a0f",
+    "validation/stages/adapters.py": "52d4823e3143fb80bc69677c3c20954f2abfdcb16f64035ad2aa0d9a6f673fba",
+    "validation/stages/container_reuse.py": "a2bf45eaa39205a930e668b5141df794ff45afebde4c0ada4b2bbb6503738e52",
+    "validation/stages/harbor.py": "0b8a8e11ba417c4617e2245cbf2962b2adae87136642fc21b82abe3afa7f44c6",
+    "validation/stages/harden_entry.py": "a49fc58f18dd8fee9af91c239ccf293e15c27f0f3d14294e49865e84920fe746",
+    "validation/stages/paired.py": "e8f855eb92ca88c66e6080d425a8fc584fc77cb8758c175b10ab48efff205c29",
+    "validation/stages/runner.py": "a67869224e9540b8de6745d496468cc99f64c6c52a44fed03674c673e04106e6",
+    "validation/stages/stage_10_hacker_fixer_loop.py": "0798bb3471f81aa85f726fa5984eb6515c52ffb4e637bd04849180819a7576e6",
+    "validation/stages/stage_1_static_checks.py": "0962f36152a8333946409d9db7580e4034d335d9a58a46d0cd72ab0a03f2908b",
+    "validation/stages/stage_2_llm_rubric_review.py": "afd4bec1b9f5f3c0cfa4269f0b9126ffff6ff1a643fdb52405ef60f620639f86",
+    "validation/stages/stage_3_build_validation.py": "40b24e4bf291754b1fdbc1381de13f9929c1b513e3de0a58813ca36f59601e9e",
+    "validation/stages/stage_4_oracle_validation.py": "d67771af6b7a39824dea6e2d42e8f403762f58eca35d420120911e34dc8c928d",
+    "validation/stages/stage_5_nop_validation.py": "287e7a7437fab8e704f3a80cc16536feddc94db4a1a345de181e065e2a9ef1e4",
+    "validation/stages/stage_6_agent_trials.py": "25662fc1900e556c3d174c17e7154c6c8793345cced547c232f4209588008182",
+    "validation/stages/stage_7_trace_metrics.py": "eecd7b111182bbd6e4325f0a781dcb7980d47bf6b09ed4fd912d0b432a2327e5",
+    "validation/stages/stage_8_trajectory_analysis.py": "4828e96dcaf1c69cccde1c71dcedb0d9ba5719fe0a4340654aa89778125515b1",
+    "validation/stages/stage_9_cheat_trials.py": "c1f01d4f10108e12358671fc27824f2c98057636caa707f40287f0a0792ffde9",
+    "validation/static_resume.py": "6c2e21c0ea0c99364d393662472f0e28a0df6e8ebb42cbc83512d386e6881d75",
+    "validation/upstream/__init__.py": "30be6817226408a29f1e752bdd4147bd1e8f386f294dfe37b2b1743008e01f2a",
+    "validation/upstream/__main__.py": "fb76f9492685add6774fdf9faf6ee0cffcd78c5860727b5c671351429b484194",
+    "validation/verify/check_bridge.py": "47a85a5f578b253ce5780b191f6f0175d4c3bb26ca2032357d9ca9d5773b2123",
+    "validation/verify/check_images.py": "51961c368707087412a6e46edc487fa9751a6a8fb19ab415abcc2cdf21a363bd",
+    "validation/verify/check_isolation.py": "3d0c5fa1f8df0d7790d2ab16acfd51270475dd811baa875bb82106bd6c812040",
+    "validation/verify/check_reproducible.py": "f501414b8a24893ce783dca97eb8ef2e516a7224fe9a6db55e1a6f7ff80747a5",
+    "validation/verify/check_terminal_bench.py": "1a1dde459c489c09f3bbd40859e1677f39ca4f09153af3828baa259d13ca9997",
+    "validation/verify/pass_at_k.py": "107240a8f7645124b2393563502cd14da656f7cf7327e042b1a22269395556d8",
+    "validation/verify/plot_pass_rates.py": "c8124a169120cdb01fcea0692c6854154f2cd996e44f12fdafb42ef61d79ca97",
+    "validation/verify/vendor/terminal_bench/scripts/checks/check_ai_detection.py": "3cdcdc9229f9be6d2d7025fb179295be2039a76a9cac2b89228a85197f15d9cb"
+  },
+  "arguments": {
+    "tasks": "/hnvme/workspace/y500bb10-termigen-audit/phase1-static-20261007-152318/full/contract.stage1-tasks",
+    "fix_instruction_suffix": true,
+    "dataset_source": null,
+    "dataset_revision": null,
+    "min_tasks": 3566,
+    "out": "/hnvme/workspace/y500bb10-termigen-audit/followup-20261007-193743/baseline/report",
+    "backend": "apptainer",
+    "environment_kwargs": {},
+    "agent": "terminus-2",
+    "review_agent": null,
+    "model": null,
+    "review_model": null,
+    "proposal_review": false,
+    "proposal_model": null,
+    "review_local": false,
+    "analysis_agent": null,
+    "analysis_model": null,
+    "agent_kwargs": {},
+    "attempts": 1,
+    "concurrency": 2,
+    "static_concurrency": null,
+    "static_resume": null,
+    "static_resume_accept_previous_path_check": false,
+    "reuse_validation_containers": true,
+    "container_start_concurrency": 2,
+    "container_start_interval": 0.25,
+    "limit": null,
+    "task_id_range": null,
+    "static_profile": "training",
+    "exclude": [],
+    "submit": "never",
+    "time": "04:00:00",
+    "partition": "auto",
+    "cpus": 48,
+    "gpus": null,
+    "memory": "32G",
+    "trial_cpus": null,
+    "trial_memory_mb": null,
+    "serve_model": null,
+    "serve_weights": null,
+    "serve_image": null,
+    "serve_context": 32768,
+    "serve_cpus": 4,
+    "serve_memory_mb": 32768,
+    "max_num_seqs": null,
+    "api_base": null,
+    "network_mode": "host",
+    "reward_key": "reward",
+    "trials": null,
+    "max_iterations": 10,
+    "timeout_minutes": 120,
+    "publish_repo": null,
+    "publish_require_complete": false,
+    "publish_folder": [],
+    "publish_readme": false,
+    "publish_readme_model": "claude-fable-5-1",
+    "publish_readme_seed": 0,
+    "publish_readme_evidence": [],
+    "publish_analysis": false,
+    "force_build": false,
+    "dry_run": false
+  },
+  "task_manifest": {
+    "path": "aggregate-contract.tasks.json",
+    "count": 3566,
+    "sha256": "e189b26e164fa58f8db5ff578a27838f90bd0250d38ced99fc2686fd7c2f389a"
+  },
+  "sha256": "b532a6c69a95fb7bfabb293ac97d353229ee5bf4067d02819376633d0e230da4"
+}

evidence/termigen-phases-1-3-5-6-7-preview-20261014/final-outcome-summary.json

@@ -0,0 +1,370 @@
+{
+  "checked_at": "2026-10-14T07:25:48.387947+00:00",
+  "model": "Qwen/Qwen3-30B-A3B-Instruct-2507",
+  "agent": "terminus-2",
+  "complete_batches": 17,
+  "tasks": 3336,
+  "attempts": 13344,
+  "canonical_results_verified": 13344,
+  "attempts_per_task": 4,
+  "pass@1": 0.3754496402877698,
+  "pass@4": 0.5545563549160671,
+  "solved_attempts": 5010,
+  "solved_tasks": 1850,
+  "graded": 13165,
+  "missing_rewards": 179,
+  "missing_traces": 32,
+  "exceptions": {
+    "AgentTimeoutError": 1816,
+    "RuntimeError": 37,
+    "TaskMemoryLimitError": 64,
+    "VerifierTimeoutError": 11,
+    "TmuxBatchProtocolError": 12,
+    "TmuxSessionEndedError": 47,
+    "AddTestsDirError": 7,
+    "TmuxCommandError": 1
+  },
+  "terminations": {
+    "task_timeout": 1816,
+    "task_complete": 11344,
+    "error": 101,
+    "task_memory_limit": 64,
+    "session_ended": 19
+  },
+  "malformed_tool_calls": 17532,
+  "trajectories_with_malformed_calls": 1088,
+  "cache": {
+    "cached": 3340
+  },
+  "gpu_sample_weighted_utilization": 99.01567991349013,
+  "resources": {
+    "start_local": "2026-10-13T14:02:14",
+    "end_local": "2026-10-13T18:29:11",
+    "elapsed_hours": 4.449166666666667,
+    "gpu_hours": 53.571666666666665,
+    "min_job_minutes": 166.53333333333333,
+    "max_job_minutes": 209.25,
+    "mean_job_minutes": 189.07647058823528,
+    "states": {
+      "FAILED": 17
+    }
+  },
+  "comparison": {
+    "previous_model": "Qwen3-Coder-30B-A3B-Instruct",
+    "previous_pass@1": 0.48456235011990406,
+    "previous_pass@4": 0.6363908872901679,
+    "pass@1_delta_pp": -10.911270983213427,
+    "pass@4_delta_pp": -8.183453237410077,
+    "previous_gpu_hours": 70.24388888888889,
+    "previous_missing_rewards": 142,
+    "previous_agent_timeouts": 414,
+    "new_only_solved_tasks": 154,
+    "old_only_solved_tasks": 427,
+    "union_solved_tasks": 2277,
+    "note": "Task scope and hashes match the prepared comparison. Model and concurrency both changed, so runtime and outcome changes are not attributable to one factor alone."
+  },
+  "batches": [
+    {
+      "batch": 1,
+      "job_id": "1004595",
+      "tasks": 197,
+      "attempts": 788,
+      "exceptions": {
+        "AgentTimeoutError": 119,
+        "RuntimeError": 4,
+        "TaskMemoryLimitError": 1,
+        "VerifierTimeoutError": 1,
+        "TmuxBatchProtocolError": 2
+      },
+      "cache": {
+        "cached": 197
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-01/submissions/c592298972c0/evidence.tar.gz"
+    },
+    {
+      "batch": 2,
+      "job_id": "1004596",
+      "tasks": 197,
+      "attempts": 788,
+      "exceptions": {
+        "AgentTimeoutError": 117,
+        "TaskMemoryLimitError": 4,
+        "TmuxSessionEndedError": 2,
+        "RuntimeError": 4,
+        "TmuxBatchProtocolError": 1
+      },
+      "cache": {
+        "cached": 198
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-02/submissions/49fb729b5bd4/evidence.tar.gz"
+    },
+    {
+      "batch": 3,
+      "job_id": "1004597",
+      "tasks": 197,
+      "attempts": 788,
+      "exceptions": {
+        "TmuxSessionEndedError": 4,
+        "AgentTimeoutError": 101,
+        "TaskMemoryLimitError": 2,
+        "AddTestsDirError": 1,
+        "RuntimeError": 4
+      },
+      "cache": {
+        "cached": 197
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-03/submissions/7000247734bf/evidence.tar.gz"
+    },
+    {
+      "batch": 4,
+      "job_id": "1004598",
+      "tasks": 197,
+      "attempts": 788,
+      "exceptions": {
+        "AgentTimeoutError": 113,
+        "VerifierTimeoutError": 1,
+        "TaskMemoryLimitError": 1,
+        "RuntimeError": 4,
+        "TmuxSessionEndedError": 1,
+        "TmuxBatchProtocolError": 1
+      },
+      "cache": {
+        "cached": 197
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-04/submissions/e076551bf489/evidence.tar.gz"
+    },
+    {
+      "batch": 5,
+      "job_id": "1004599",
+      "tasks": 196,
+      "attempts": 784,
+      "exceptions": {
+        "AgentTimeoutError": 94,
+        "TaskMemoryLimitError": 1,
+        "TmuxSessionEndedError": 1
+      },
+      "cache": {
+        "cached": 196
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-05/submissions/8cec8babc1a9/evidence.tar.gz"
+    },
+    {
+      "batch": 6,
+      "job_id": "1004600",
+      "tasks": 196,
+      "attempts": 784,
+      "exceptions": {
+        "AgentTimeoutError": 110,
+        "TmuxSessionEndedError": 5,
+        "TaskMemoryLimitError": 12,
+        "RuntimeError": 1,
+        "AddTestsDirError": 3,
+        "VerifierTimeoutError": 1
+      },
+      "cache": {
+        "cached": 196
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-06/submissions/a402558dba2e/evidence.tar.gz"
+    },
+    {
+      "batch": 7,
+      "job_id": "1004601",
+      "tasks": 196,
+      "attempts": 784,
+      "exceptions": {
+        "AgentTimeoutError": 113,
+        "TaskMemoryLimitError": 4,
+        "RuntimeError": 4,
+        "AddTestsDirError": 1,
+        "TmuxSessionEndedError": 2
+      },
+      "cache": {
+        "cached": 196
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-07/submissions/808e2eb233e7/evidence.tar.gz"
+    },
+    {
+      "batch": 8,
+      "job_id": "1004602",
+      "tasks": 196,
+      "attempts": 784,
+      "exceptions": {
+        "AgentTimeoutError": 101,
+        "TaskMemoryLimitError": 13,
+        "TmuxSessionEndedError": 2,
+        "VerifierTimeoutError": 3,
+        "AddTestsDirError": 1
+      },
+      "cache": {
+        "cached": 196
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-08/submissions/adc589f99ecb/evidence.tar.gz"
+    },
+    {
+      "batch": 9,
+      "job_id": "1004603",
+      "tasks": 196,
+      "attempts": 784,
+      "exceptions": {
+        "AgentTimeoutError": 106,
+        "TmuxSessionEndedError": 7,
+        "TaskMemoryLimitError": 4,
+        "RuntimeError": 4
+      },
+      "cache": {
+        "cached": 196
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-09/submissions/6b59fd8d7c8e/evidence.tar.gz"
+    },
+    {
+      "batch": 10,
+      "job_id": "1004604",
+      "tasks": 196,
+      "attempts": 784,
+      "exceptions": {
+        "AgentTimeoutError": 106,
+        "TmuxSessionEndedError": 2,
+        "TmuxBatchProtocolError": 1,
+        "TaskMemoryLimitError": 1
+      },
+      "cache": {
+        "cached": 196
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-10/submissions/d9f4d29e7d9b/evidence.tar.gz"
+    },
+    {
+      "batch": 11,
+      "job_id": "1004605",
+      "tasks": 196,
+      "attempts": 784,
+      "exceptions": {
+        "AgentTimeoutError": 100,
+        "TmuxSessionEndedError": 3,
+        "TaskMemoryLimitError": 4,
+        "VerifierTimeoutError": 1
+      },
+      "cache": {
+        "cached": 196
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-11/submissions/b827c7005df3/evidence.tar.gz"
+    },
+    {
+      "batch": 12,
+      "job_id": "1004606",
+      "tasks": 196,
+      "attempts": 784,
+      "exceptions": {
+        "AgentTimeoutError": 130,
+        "TmuxBatchProtocolError": 4,
+        "TmuxSessionEndedError": 2,
+        "TaskMemoryLimitError": 1
+      },
+      "cache": {
+        "cached": 196
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-12/submissions/29c7d1b08767/evidence.tar.gz"
+    },
+    {
+      "batch": 13,
+      "job_id": "1004607",
+      "tasks": 196,
+      "attempts": 784,
+      "exceptions": {
+        "TmuxSessionEndedError": 3,
+        "AgentTimeoutError": 99,
+        "TaskMemoryLimitError": 3,
+        "TmuxBatchProtocolError": 1,
+        "AddTestsDirError": 1
+      },
+      "cache": {
+        "cached": 198
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-13/submissions/098b64b83ce4/evidence.tar.gz"
+    },
+    {
+      "batch": 14,
+      "job_id": "1004608",
+      "tasks": 196,
+      "attempts": 784,
+      "exceptions": {
+        "TmuxBatchProtocolError": 1,
+        "AgentTimeoutError": 110,
+        "TmuxSessionEndedError": 6,
+        "TaskMemoryLimitError": 2,
+        "TmuxCommandError": 1,
+        "VerifierTimeoutError": 1
+      },
+      "cache": {
+        "cached": 196
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-14/submissions/980d12c03fdb/evidence.tar.gz"
+    },
+    {
+      "batch": 15,
+      "job_id": "1004609",
+      "tasks": 196,
+      "attempts": 784,
+      "exceptions": {
+        "AgentTimeoutError": 87,
+        "RuntimeError": 5,
+        "TaskMemoryLimitError": 2
+      },
+      "cache": {
+        "cached": 196
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-15/submissions/bd5f20344c1b/evidence.tar.gz"
+    },
+    {
+      "batch": 16,
+      "job_id": "1004610",
+      "tasks": 196,
+      "attempts": 784,
+      "exceptions": {
+        "AgentTimeoutError": 101,
+        "TmuxSessionEndedError": 5,
+        "RuntimeError": 5,
+        "VerifierTimeoutError": 2,
+        "TaskMemoryLimitError": 3,
+        "TmuxBatchProtocolError": 1
+      },
+      "cache": {
+        "cached": 197
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-16/submissions/3a9a96d016cc/evidence.tar.gz"
+    },
+    {
+      "batch": 17,
+      "job_id": "1004611",
+      "tasks": 196,
+      "attempts": 784,
+      "exceptions": {
+        "TaskMemoryLimitError": 6,
+        "AgentTimeoutError": 109,
+        "TmuxSessionEndedError": 2,
+        "RuntimeError": 2,
+        "VerifierTimeoutError": 1
+      },
+      "cache": {
+        "cached": 196
+      },
+      "complete": true,
+      "archive": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-17/submissions/3e82cf7e1cf7/evidence.tar.gz"
+    }
+  ]
+}

evidence/termigen-phases-1-3-5-6-7-preview-20261014/input-provenance.json

@@ -0,0 +1,343 @@
+{
+  "validation_contract": {
+    "path": "/hnvme/workspace/y500bb10-termigen-audit/followup-20261007-193743/aggregate-contract.json",
+    "sha256": "674760956c8858ec6b8965ff3b1fab5fdc9e19c3991838fb3c32379569229104"
+  },
+  "gate_reports": [
+    {
+      "path": "/hnvme/workspace/y500bb10-termigen-audit/followup-20261007-193743/merged/report/stage-1.json",
+      "sha256": "1bbbc1bad844a8988fe905540f262b7909ef55fab490fdae4f03924ac69abe02"
+    },
+    {
+      "path": "/hnvme/workspace/y500bb10-termigen-audit/followup-20261007-193743/merged/report/stage-5.json",
+      "sha256": "61b3e02b859edda6a387c7866936a65e6326f99f59beea1dfff912de2affb8be"
+    },
+    {
+      "path": "/hnvme/workspace/y500bb10-termigen-audit/publish-phase3-reviewed-20261009-113141/report/stage-3-reviewed.json",
+      "sha256": "944e83c71fdc5b483484c2f34eb3505ccccd1748311f585879070a193e54b451"
+    }
+  ],
+  "teacher": {
+    "kind": "aggregation of completed informational teacher reports",
+    "source_contracts_preserved": true,
+    "task_hashes_verified_against_validation_contract": 3336,
+    "sources": [
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-01.json",
+          "sha256": "05a82036c60c0eed1b26a5f7cd19323034275b6b2f4db80d7e648ac919bed6bc"
+        },
+        "contract_sha256": "0f0a6db977655eb3ab25072fdc563357394aa6300c8b83837971d8a4ce1299ff",
+        "job_id": "1004595",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-01/submissions/c592298972c0/report/stage-6-ba90bb456867.json",
+            "sha256": "f609cbc8f5f9076c6948f023668bea81ef16a6c1b1465aa5305abd2c2842d86e"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-01/submissions/c592298972c0/report/stage-7-bbaf19d5da1e.json",
+            "sha256": "d835f8fa856fade22c72b4efb6f6381d0b3cea055ef15027e08546d10375e600"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-02.json",
+          "sha256": "0d65cf3eac3c496b96a42579b86119372bc49269c06556c6a3fac2f0acc5cee2"
+        },
+        "contract_sha256": "9ea31ce25d37ed7892378f7c8719aa946e44a6bf5e22736cd7e41a7e6f2159ed",
+        "job_id": "1004596",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-02/submissions/49fb729b5bd4/report/stage-6-408d600edcf3.json",
+            "sha256": "4be054d31313a34e5c5f61544155070911577065292f741c46c8429652a87130"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-02/submissions/49fb729b5bd4/report/stage-7-392521b77b22.json",
+            "sha256": "2393847b2af18685452c71a3c1ac5d733923b2a3ff8c6ca667656d05376c7cd4"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-03.json",
+          "sha256": "374836aed98a0509c3418f2f88d6717d1bdd5297bf19b41a93ae901d4fe738fd"
+        },
+        "contract_sha256": "eff36135da7c4f09b475861e55c42f46ac663fbe8e6af6510111afb4497f8820",
+        "job_id": "1004597",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-03/submissions/7000247734bf/report/stage-6-6bf04dc6d5d8.json",
+            "sha256": "c37e9c18fbbcc9041270a0c47c423f0f9d8c2562adf86267cadc5b8b8dd0e26f"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-03/submissions/7000247734bf/report/stage-7-0a8725473fdf.json",
+            "sha256": "64c370c89ab47274cd011f16a4f7457a15e68426204839fcebbb95c38cbe01c8"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-04.json",
+          "sha256": "ff99d070e20c28f066d3bb6111850e18653ad08aef33ef305f219bb0625c4bef"
+        },
+        "contract_sha256": "c7bcbb0e62834f211bb9128de2f778db6f067c7dc30ce2a5f5d0a636eb970475",
+        "job_id": "1004598",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-04/submissions/e076551bf489/report/stage-6-830c7baf4668.json",
+            "sha256": "f3072130d8754ce5e942251fc93b2449991e6dace605c018a0859bd7ed73f179"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-04/submissions/e076551bf489/report/stage-7-3b428dfda0a4.json",
+            "sha256": "8333fcf843d922639546204779035372e5aa170e6bc49a42097ef8462e9877ca"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-05.json",
+          "sha256": "5ddf373804d071d6a0afe325b388bb1ad10705d94bb40cb440a0bffc30470fc4"
+        },
+        "contract_sha256": "998d1ec40aba845dae34211fc18188c9d1e5d5481c54869c375b371ac8f7cbb3",
+        "job_id": "1004599",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-05/submissions/8cec8babc1a9/report/stage-6-cbb149e9f05c.json",
+            "sha256": "b07ccd8168bbbd1a581817b593077a24fd7b4509dcd94df690a763f8b7b47201"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-05/submissions/8cec8babc1a9/report/stage-7-5ed70b93f71c.json",
+            "sha256": "d0b243936f6fa26ee5a533d72ae03b136f852b91f62a986882b26aa1a3d3938b"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-06.json",
+          "sha256": "3fadd90062e6cbc02f42a8d4e08b1cc174526b3aacdaf70ee1684d22ca107e51"
+        },
+        "contract_sha256": "0a0ad87c61b1a7b41ae5fbef064de8a0987d30859393aaf05d3c19c903db2acc",
+        "job_id": "1004600",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-06/submissions/a402558dba2e/report/stage-6-86f471d45257.json",
+            "sha256": "4bd34867eb7169d29dd22dc8a48d20bdd78fbb8e272da63c8f03cca5da699d1e"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-06/submissions/a402558dba2e/report/stage-7-5982b4e7bbbf.json",
+            "sha256": "163cb6c64d571051d6741f749814f43563a35ddd1b20bc9955b0429a5862cab5"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-07.json",
+          "sha256": "5b02c5673580d9e5cee4570a51c3fcefe83df83fe6c39df5c8591aa6966da5dd"
+        },
+        "contract_sha256": "ef01983cf688d880553839eeed4c8a82f25d6cea2f9b93ba211a187ce6be4ccb",
+        "job_id": "1004601",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-07/submissions/808e2eb233e7/report/stage-6-e52efa12f494.json",
+            "sha256": "a798211ed6e8a52944a3cff837b4e9e5d74057606ebb035502355d802a9fc586"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-07/submissions/808e2eb233e7/report/stage-7-fd48ea1a2f43.json",
+            "sha256": "d4273d3718fb22423da2e3d621042eabb85c57873c79802e7f60b3822a07095c"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-08.json",
+          "sha256": "7a9b38b04a718dde8a3ce6e5be85d4ed06b03e478db174fe6e67e1cf780b5168"
+        },
+        "contract_sha256": "dc5a7a99a8c67bfff09a767dbf41e243b80510fd00a275b8cdabf6a3bdf04e1c",
+        "job_id": "1004602",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-08/submissions/adc589f99ecb/report/stage-6-56d0fa2f32e8.json",
+            "sha256": "7380ec1c420efaf7badf49a995ad7e82c4566ac25fb8d265da170210a29b7e88"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-08/submissions/adc589f99ecb/report/stage-7-2686df7e06fe.json",
+            "sha256": "9e65ca652aabbe7395524eb48b6d62f7002280fbb4ec2ac7094aa895b576f3ef"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-09.json",
+          "sha256": "9fe5dd82a926a729bd5c11c5145bf409eff4de51b0f6a784e070f0ddfcee9a82"
+        },
+        "contract_sha256": "10a5b50e89173c02c6c55b06e0f6f07990682385f9f0fa1ffcc74b1d8e0b5bef",
+        "job_id": "1004603",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-09/submissions/6b59fd8d7c8e/report/stage-6-58da70d1212c.json",
+            "sha256": "fd279ee130b79685effdc75aa27ab4424f2f18b23c30242d7d8c03df38624b0a"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-09/submissions/6b59fd8d7c8e/report/stage-7-85616c319c1f.json",
+            "sha256": "fd1a3f8ff49b1c2fc9194c3ce0271b3b1d8ab738aa84725b6774d4f7deac25f1"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-10.json",
+          "sha256": "7e84c1626e57f4420128929a78eee97f171fa010c3c26ae102da0ca41bd4c83b"
+        },
+        "contract_sha256": "fab5fea81a32ddb43b71c90ce6dfa69f67eed40bb3d5f524a1365c3cc0916e8a",
+        "job_id": "1004604",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-10/submissions/d9f4d29e7d9b/report/stage-6-3890e2429a2b.json",
+            "sha256": "c1933b9e58cafea347ab96788af0b6e50031d3ca0cfbe909cac2531d113b267e"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-10/submissions/d9f4d29e7d9b/report/stage-7-2fed5abb8d38.json",
+            "sha256": "7d65abe1005a6009793dd1a8e968c370db04fa65b730a721edd80f2e6ef63c44"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-11.json",
+          "sha256": "42f9d36cdeb9f4a541d5efc4669e79a15129e27bdc03850a5efb37bd336e7be6"
+        },
+        "contract_sha256": "78c80cee0a49ec41b7e317c84bcd6c6030a336b2c470f0334f9b7ad013ec1c8f",
+        "job_id": "1004605",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-11/submissions/b827c7005df3/report/stage-6-a7eef5bfa6a3.json",
+            "sha256": "88f8b7c5e8e1ca7d86ca582bb8ac18af7a0c44e04c8da0f57f0405945675da15"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-11/submissions/b827c7005df3/report/stage-7-614e55600be5.json",
+            "sha256": "7bd5b34d4bf7754800f7306daa27cbb521aa18daf3782136b36a9e2c2b5170bd"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-12.json",
+          "sha256": "d0dc52096816eba0e4e19089422dbec36b80d546816644ddc9b7d1e5933e8d28"
+        },
+        "contract_sha256": "dd7777b1dcdf2370007d32c477cc24425deefd03c3440d72f54bd617cc0acc8b",
+        "job_id": "1004606",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-12/submissions/29c7d1b08767/report/stage-6-357edc4ad3d5.json",
+            "sha256": "46ad7789c7fa70968a40d8fc5bed47fd33bd9694fe57032667cfb02d00ca72ea"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-12/submissions/29c7d1b08767/report/stage-7-ffc97971197a.json",
+            "sha256": "8aa996ad1a0eac4c95c550aef3e494e7415c7fd9e755f3e931b3b17faaf0a1b4"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-13.json",
+          "sha256": "55cf356cf4ced66bcca5a1e5a3e697794bb44235a052dfbc3e08c542c3a625c1"
+        },
+        "contract_sha256": "cb7b5b687b3eddbbeaafe1adae207102e2af2de060435fcb59b07af75fe68926",
+        "job_id": "1004607",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-13/submissions/098b64b83ce4/report/stage-6-732c576c8c35.json",
+            "sha256": "8bc11b6071aa4b69ea23db22db89a01964b55028e268e9854706de7401e01720"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-13/submissions/098b64b83ce4/report/stage-7-b41795029df8.json",
+            "sha256": "124bb40c8340b6223a941878bd8df705486bbe49139588ef6b7a7f048b38b064"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-14.json",
+          "sha256": "5c958de913c516ff3d35662faddedc55c72a9850ae018eb594f849f033733a5f"
+        },
+        "contract_sha256": "37d0f439b90387a81abb5d8932ff8b74c55d6f418bef9c02d483ce39a7c5ecf6",
+        "job_id": "1004608",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-14/submissions/980d12c03fdb/report/stage-6-0074de82f377.json",
+            "sha256": "ae713c33a6e346e479afc56162d1a4ed56a852785632bc49829a1ff698103264"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-14/submissions/980d12c03fdb/report/stage-7-e6ff0cd5975f.json",
+            "sha256": "97d774cc0b1887c477d6c0d95593f2c4ea438fef552d6763c13ef3aa05ec1ac3"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-15.json",
+          "sha256": "4e7d89a3c9bfba3e84bbe4f0d41b1dd6105c5b1f8b29fe0e3fbec976723a75f9"
+        },
+        "contract_sha256": "3599881669b946c955c7f8df65468652fbe6a3b3acd32b03f8cfd88e40697461",
+        "job_id": "1004609",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-15/submissions/bd5f20344c1b/report/stage-6-774d30401fd5.json",
+            "sha256": "b7378ed3af3a21bcc682dad01ceb3be4861bce5e3cb9bda94b6b7db30d6dc3c6"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-15/submissions/bd5f20344c1b/report/stage-7-0f63e6ab1863.json",
+            "sha256": "4c062e1d96143c9297a61a460fe40f969324ffc982627f410a8503315b32f3d6"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-16.json",
+          "sha256": "8ac6a6d596614e3877543a1ba7dc0e759d18f28bb9663c09ed8633e0050e022b"
+        },
+        "contract_sha256": "79a5c9d111da8352f3da30353bd8f27df12275dc3ddf527be6ab2128bff163cc",
+        "job_id": "1004610",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-16/submissions/3a9a96d016cc/report/stage-6-d095fe78f493.json",
+            "sha256": "097acc68ead35ad7ccb02cb8206de73edb42f41d877bd50e24ad4f5def0ecbe8"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-16/submissions/3a9a96d016cc/report/stage-7-abd78587bdc4.json",
+            "sha256": "c8befc6acfe6a6e8f5c84031faf43f22976cf2f765fe5678ae75c489418c22b4"
+          }
+        }
+      },
+      {
+        "contract": {
+          "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/contracts/batch-17.json",
+          "sha256": "906c3853fcd0da392a9ea6cdb059c7ffd4ac9b23d87dbde7f8ab09378ca1a903"
+        },
+        "contract_sha256": "9c3b6c6c2ab51fd5d6ec096ec23f426d6da5915002272e84641f6f75c928994c",
+        "job_id": "1004611",
+        "reports": {
+          "6": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-17/submissions/3e82cf7e1cf7/report/stage-6-ad28db06670d.json",
+            "sha256": "b30905483ebbceb92b7a3b6a0fc8a0102634883437a85c7d944b0396366e4599"
+          },
+          "7": {
+            "path": "/hnvme/workspace/y500bb10-termigen-audit/teacher-qwen3-30b-instruct-2507-20261013/full-prepared-17jobs-20261013/runs/batch-17/submissions/3e82cf7e1cf7/report/stage-7-6e87f1aefc88.json",
+            "sha256": "652634e28f54498d991636a11d6bc11d4908a4ff300d10a50966d5fdaf993027"
+          }
+        }
+      }
+    ],
+    "note": "The validation contract covers phases 1,3,5. Each teacher batch has its own frozen contract. These derived reports do not claim execution under the validation contract."
+  },
+  "skipped_stages": {
+    "4": "TermiGen upstream tasks have no reference solutions."
+  },
+  "not_required_checks": [
+    "check-pip-pinning.sh",
+    "check-test-file-references.sh"
+  ],
+  "all_payload_hashes_verified": 3566,
+  "teacher_inclusion_changes_filtering": false
+}

evidence/termigen-phases-1-3-5-6-7-preview-20261014/patcher-reproduction.json

@@ -0,0 +1,7 @@
+{
+  "patcher": "/home/hpc/y500bb/y500bb10/repos/OT-next-data/data/termigen/patch.py",
+  "patcher_sha256": "a3d8411b8712cc46c560196d16be8f16b2cd7be0ed7c2098256dc150224d1593",
+  "tasks": 3566,
+  "binary_mismatches": [],
+  "content_mismatches": []
+}

evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-1.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1bbbc1bad844a8988fe905540f262b7909ef55fab490fdae4f03924ac69abe02
+size 19151774

evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-3.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:944e83c71fdc5b483484c2f34eb3505ccccd1748311f585879070a193e54b451
+size 15320434

evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-7.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:913048363bef7371b7f0f8b194b83e61858c1e36c060c9cf46fb94253dbd5bb0
+size 25559007

evidence/termigen-phases-1-3-5-6-7-preview-20261014/teacher-comparison.json

@@ -0,0 +1,40 @@
+{
+  "models": [
+    {
+      "model": "Qwen3-Coder-30B-A3B-Instruct",
+      "agent": "terminus-2",
+      "tasks": 3336,
+      "attempts": 13344,
+      "pass@1": 0.48456235011990406,
+      "pass@2": 0.5742406075139888,
+      "pass@3": 0.6137589928057554,
+      "pass@4": 0.6363908872901679,
+      "solved_tasks": 2123,
+      "missing_rewards": 142,
+      "missing_traces": 32,
+      "agent_timeouts": 414,
+      "source_report": "/hnvme/workspace/y500bb10-termigen-audit/teacher-full-20261008/stage-6-aggregate.json",
+      "source_sha256": "7d104fd9e9fe48455debf18dd11c56bb7e3134f59c1c969682dc346c818e5878"
+    },
+    {
+      "model": "Qwen3-30B-A3B-Instruct-2507",
+      "agent": "terminus-2",
+      "tasks": 3336,
+      "attempts": 13344,
+      "pass@1": 0.3754496402877698,
+      "pass@2": 0.4741207034372502,
+      "pass@3": 0.5233063549160671,
+      "pass@4": 0.5545563549160671,
+      "solved_tasks": 1850,
+      "missing_rewards": 179,
+      "missing_traces": 32,
+      "agent_timeouts": 1816,
+      "source_report": "/hnvme/workspace/y500bb10-termigen-audit/publish-phases-1-3-5-6-7-dry-run-20261014/report/stage-6.json",
+      "source_sha256": "8c7f225292d0d1aa647d5afe85e33d7d2f7607c2d15c0ec4009a41805854a9ab"
+    }
+  ],
+  "filtering": false,
+  "denominator": "Same 3336 validated tasks, four recorded attempts per task and model. Missing rewards count as unsolved only for pass-rate calculations; stored missing values remain missing.",
+  "estimator": "Mean over tasks of 1 - C(4-solved,k)/C(4,k).",
+  "comparison_limit": "Both model and concurrency differ (Coder 10, Instruct 16)."
+}

pull-request.md

@@ -0,0 +1,96 @@
+**No gold/reference-solution checks were run: TermiGen provides no gold solutions. Broken tasks will likely remain in this dataset and will be filtered out in the next version.** Teacher passes are evidence of verifier acceptance, not proof that a task or its tests are correct.
+
+## TermiGen
+
+Keeps **3,369 tasks** and archives **197** from [ucsb-mlsec/terminal-bench-env](https://github.com/ucsb-mlsec/terminal-bench-env/tree/03bddac74ada2fa344df0e93b4a3ace2034294d1), covering all **3,566 source tasks**. This PR includes the task payloads, archived tasks with their reasons, and the supporting validation and teacher reports.
+
+The earlier [stage-3 publication (PR #8)](https://huggingface.co/datasets/FWeindel/validated-tasks/discussions/8) retained 3,506 tasks and archived 60. This PR adds static and no-op results, excluding a further **137 tasks**: 133 with absolute-path findings, 2 with CPU-count findings, and 2 whose verifier passed with no answer. No previously archived task is restored.
+
+Filtering starts with all 3,566 source tasks and applies stages 1, 3 and 5 in order. Each excluded task is counted once, at its first applicable exclusion. The counts use the full validation run and recorded environment reruns.
+
+| Check | Tasks entering this step | Newly archived | Tasks retained after this step |
+| --- | ---: | ---: | ---: |
+| 1: Static checks, with the recorded advisory exceptions | 3,566 | 141 | 3,425 |
+| 3: Environment build, startup and inspected files | 3,425 | 54 | 3,371 |
+| 5: No-op verification | 3,371 | 2 | 3,369 |
+
+**Only 3,336 retained tasks passed all three measured validation phases.** The remaining 33 are retained as unresolved under the publisher's existing policy: 28 verifiers executed no tests and 5 verifiers timed out. These are not passing no-op results. Stage 4 was not run because there are no reference solutions.
+
+Static results use the recorded training-profile exclusions; pip-pinning and test-file-reference findings remain advisory. The detailed policy and per-task findings are preserved in the [run record](runs/termigen-phases-1-3-5-6-7-preview-20261014.json).
+
+## Archived tasks
+
+The first applicable exclusion accounts for each task once: **141 at stage 1, 54 at stage 3, and 2 at stage 5**. Stage 3 still has the same 60 archived outcomes as PR #8; six of those tasks are counted at stage 1 in this combined report.
+
+| Archive label | Tasks | Explanation |
+| --- | ---: | --- |
+| build-command-unavailable | 1 | The build could not find a required shell command. |
+| container-start-failed | 7 | Apptainer could not start the environment; the saved logs include fakeroot startup diagnostics. |
+| copy-destination-collision | 2 | Build-context copying collided with a file already present at the destination. |
+| dependency-build-failed | 12 | Environment setup or dependency installation failed. Per-task errors are preserved in the linked stage report. |
+| environment-check-incomplete | 9 | Environment startup or inspection did not finish successfully; the report contains no error text. These tasks remain archived under the recorded phase 3 policy. |
+| image-preparation-failed | 3 | CUDA, Lean and Ray image builds failed with dpkg package-configuration errors in all three reruns. Later attempts encountered missing images and did not rebuild them. |
+| missing-copy-source | 6 | The Dockerfile references a build-context file that is absent from the task payload. |
+| no-op-reward-one | 2 | no-answer reward [1.0] instead of 0 |
+| runtime-file-tree-mismatch | 12 | Expected task files were missing or differed from their recorded contents inside the running environment. |
+| static:compose-host-binds | 3 | static checks failed: check-compose-host-binds.sh |
+| static:nproc | 3 | static checks failed: check-nproc.sh |
+| static:task-absolute-path | 135 | static checks failed: check-task-absolute-path.sh |
+| unsupported-compose-sidecars | 2 | The Apptainer backend cannot run the required Docker Compose sidecars; all three reruns remained blocked. |
+
+The eight previously unresolved environment tasks remain archived after their recorded reruns: five require unsupported Compose sidecars and three failed image preparation. Later attempts on those three encountered missing images rather than rebuilding them. Nine other environment checks have incomplete startup or inspection evidence and remain archived under the reviewed stage-3 policy. These outcomes do not establish whether a valid solution exists.
+
+## Repairs
+
+Our patcher modified **2,219 of the original 3,566 tasks**. A task can have multiple changes, so the counts below overlap.
+
+| Change | Tasks affected | Explanation |
+| --- | ---: | --- |
+| deferred-helper | 4 | Reconstruct supplied build helpers when deferred installation runs before COPY payloads are available. |
+| env-expansion | 51 | Resolve supported Docker ENV substitutions so inherited command paths remain usable. |
+| initial-file | 3 | Materialize initial files from supplied payloads so they remain visible after container mounts. |
+| tmp-fixture | 158 | Relocate supplied fixtures hidden by temporary-directory mounts and update their path references. |
+| wildcard-copy | 1 | Expand build-context wildcard copies into explicit file copies. |
+| workspace-copy | 2,058 | Repair COPY destinations and preserve renamed files through byte-identical aliases. |
+
+The patcher retained all source tasks, including 11 for which no safe automatic patch was available. Their patch predictions and validation outcomes remain recorded separately.
+
+## Teacher results: pass@k
+
+Both models used **Terminus 2** on the same **3,336 tasks**, with **four attempts per task and model** (13,344 attempts each; 26,688 total). Stages 6 and 7 are informational and do not exclude tasks. The 33 tasks with unresolved no-op results were not included in teacher generation.
+
+| Model | Pass@1 | Pass@2 | Pass@3 | Pass@4 | Tasks solved at least once |
+| --- | ---: | ---: | ---: | ---: | ---: |
+| Qwen3-Coder-30B-A3B-Instruct | 48.46% | 57.42% | 61.38% | 63.64% | 2,123 / 3,336 |
+| Qwen3-30B-A3B-Instruct-2507 | 37.54% | 47.41% | 52.33% | 55.46% | 1,850 / 3,336 |
+
+Pass@k estimates success within k attempts from the four recorded attempts per task. Missing rewards count as unsolved in these rates; the underlying missing rewards remain missing. At k=4, the rate is the fraction of tasks solved at least once. The models solved **2,277 distinct tasks** between them, including 154 solved only by the Instruct model; that combined coverage is not a single-model pass@k result.
+
+| Model | Agent timeouts | Missing rewards | Missing traces |
+| --- | ---: | ---: | ---: |
+| Qwen3-Coder-30B-A3B-Instruct | 414 | 142 | 32 |
+| Qwen3-30B-A3B-Instruct-2507 | 1,816 | 179 | 32 |
+
+All planned result records are present, but some rewards and traces are missing. Timeout and execution-error findings are included in the results. The Instruct run completed all 17 stage-6/stage-7 batches within their 4-hour limits. Its Slurm failure statuses reflect recorded trial findings, not interrupted dataset coverage.
+
+## Runtime and images
+
+Environment validation used Apptainer on Helma. The original validation evidence records Python 3.12.7 and Harbor 0.8.1; the later eight-task environment reruns used Python 3.12.14, Apptainer 1.5.4, 48 CPUs, 64 GiB per job and concurrency 2. These are historical runs, not a fresh full-dataset reproduction.
+
+The Instruct teacher run used one H200 and 32 CPUs per job, concurrency 16, a 65,536-token context and a 16,384-token reply limit. Jobs took 2h47m to 3h29m and consumed 53.57 GPU-hours in total. The Coder run used concurrency 10 and consumed 70.24 GPU-hours; model and concurrency both changed, so this is not an isolated concurrency-speedup comparison.
+
+Existing cached images were reused for teacher generation. **No cached image bundles are included in this PR** because the legacy cache lacks the release provenance required by the current publisher. Task payloads and Dockerfiles are included.
+
+## Files and evidence
+
+The files below are included in this PR; the patch-script link points to the recorded GitHub commit. The teacher reports contain summaries and per-attempt metrics; the raw conversations and trial logs remain in the original run archives referenced by the provenance files. The image manifest records missing release bundles; it does not contain runnable images.
+
+- [Retained tasks](termigen/tasks.parquet) and [archived tasks with reasons](termigen/archive.parquet).
+- [Run record and per-task findings](runs/termigen-phases-1-3-5-6-7-preview-20261014.json).
+- Validation reports: [stage 1](evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-1.json), [reviewed stage 3](evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-3.json), and [stage 5](evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-5.json).
+- [Teacher comparison and pass@1 through pass@4](evidence/termigen-phases-1-3-5-6-7-preview-20261014/teacher-comparison.json).
+- Instruct results: [stage 6](evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-6.json), [stage 7](evidence/termigen-phases-1-3-5-6-7-preview-20261014/stage-7.json), and [run summary](evidence/termigen-phases-1-3-5-6-7-preview-20261014/final-outcome-summary.json).
+- [Coder outcome and trace summary](evidence/termigen-phases-1-3-5-6-7-preview-20261014/coder-outcome-summary.json).
+- [Input evidence and teacher-batch provenance](evidence/termigen-phases-1-3-5-6-7-preview-20261014/input-provenance.json).
+- [Patch comparison with upstream](evidence/termigen-phases-1-3-5-6-7-preview-20261014/patch-manifest.json), [patcher reproduction check](evidence/termigen-phases-1-3-5-6-7-preview-20261014/patcher-reproduction.json), and [patch script at the recorded commit](https://github.com/MLI-lab/OT-next-data/blob/f06623d236ee24515ef479446e029f79463559e3/data/termigen/patch.py).
+- [Image release manifest](images/manifests/termigen-phases-1-3-5-6-7-preview-20261014.json).

termigen/archive.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:932a272b5446a7dc7f8ba597fae7c1ad02a3ba1032e35fa1d76fadbeacbc1ae9
+size 2764655

termigen/tasks.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6477b86ffa1e637ee5f7e95507211f4999a056666d234c8bdd8adcd130ed535
+size 407115768