AI Product Market refresh (2026-09-24 23:26 UTC)

market/README.md

@@ -0,0 +1,50 @@
+# CyberMax AI Product Market
+
+A machine-readable catalog of every live CyberMax capability, so another AI (or a developer) can decide "this solves my
+problem" and call it without reading a web page. Generated, never hand-edited:
+
+```bash
+node scripts/market/build.mjs                    # fetches live status (Apify Store, Hugging Face, MCP Registry), writes market/
+node scripts/market/build.mjs --offline          # rebuild from market/live-snapshot.json
+cd scripts/market && npm install && node validate.mjs && node picker-test.mjs   # see TESTS.md
+HF_TOKEN=... python3 scripts/market/publish_hub.py   # publish (needs huggingface_hub)
+```
+
+## Where it lives
+
+| Now (live) | Later (once the owner connects Cloudflare) |
+|---|---|
+| https://shaw276-cybermax.static.hf.space/market/llms.txt (CyberMax hub Space `shaw276/cybermax`, sub-folder `market/`) | `https://<cybermax domain>/market/` on Cloudflare Pages (same files) |
+| https://shaw276-cybermax.static.hf.space/.well-known/agent-card.json | `https://<cybermax domain>/.well-known/agent-card.json` and `/llms.txt` at the site root |
+
+To move: change `HUB`/`MARKET_URL` at the top of `scripts/market/build.mjs`, rebuild, and upload `market/` to the site
+(the Cloudflare Pages deploy of the CyberMax website should copy `market/` to `/market/` and
+`market/.well-known/agent-card.json` to `/.well-known/agent-card.json`). Keep the HF copy as a mirror.
+
+## Files
+
+| File | What it is |
+|---|---|
+| `catalog.json` | Every capability: id, kind (`api_tool` / `dataset`), status (`live` / `coming_soon`), description written for agents, capabilities, use-when / not-for, inputs + outputs as JSON Schema (2020-12) with a real output sample where we have one, HTTP API, MCP (hosted tool name + local wrapper), docs, pay-per-event pricing, payment methods, examples, integration snippets (curl, Python, JavaScript, MCP config), usage notes, public stats. Schema: `schemas/catalog.schema.json`. |
+| `llms.txt`, `llms-full.txt` | [llmstxt.org](https://llmstxt.org/) v2 summary (links to `tools/*.md`) and the full inline version |
+| `tools/<name>.md` | One Markdown page per tool, the LLM-friendly detail an llms.txt link points to |
+| `openapi/<name>.json` | OpenAPI 3.1 per live Actor (Apify `run-sync-get-dataset-items` + async run), `openapi/datasets.json` for the free dataset API |
+| `mcp/<name>.tools.json` | MCP `tools/list`-shaped manifest (MCP 2025-06-18) per Actor (hosted Apify MCP tool) and for the free datasets server |
+| `agents/<name>.agent-card.json`, `.well-known/agent-card.json` | A2A v1.0 AgentCard per live tool and one for all of CyberMax |
+| `live-snapshot.json` | The live data the last build used (public numbers only) |
+| `tests/picker-tasks.json`, `TESTS.md` | Tool-picker test tasks and all validation results |
+
+## How a product joins
+
+Automatically, when its folder has `.actor/actor.json`, a `README.md` and pricing (`pricing-config.json`, or live
+pricing on the Store). Status flips from `coming_soon` to `live` as soon as the Apify Store lists it. Everything is read
+from the product's own files: title/description (`store/seo.json`, `.actor/actor.json`), input schema, dataset schema
+(declare `fields` for a typed output schema; otherwise it is inferred from a real sample in `examples/`), README
+"Ready-made inputs" presets (the examples), the MCP wrapper in `mcp/<x>-mcp/` (its tool description becomes the
+agent-facing description, and its "Use it ..." / "... instead" sentences become use-when / not-for), the MCP Registry
+entry and the HF demo/datasets (from `scripts/hf-funnel.py`). Optional `products/<name>/market.json` overrides:
+`summary`, `agentDescription`, `capabilities`, `useWhen`, `notFor`, `keywords`, `usageNotes`, `humanUrls`,
+`hfSpace`, `hfDatasets`, `related`.
+
+New dataset on Hugging Face under the CyberMax account → it joins on the next build (and the free
+`mcp/cybermax-datasets-mcp` server picks it up from the live catalog without a code change).

market/TESTS.md

@@ -0,0 +1,79 @@
+# AI Product Market: tests
+
+How `market/` is validated, and the results of the first run (2026-09-24, session_01DZyJd5THstKUP667AcankZ). Re-run after
+every `node scripts/market/build.mjs`:
+
+```bash
+cd scripts/market && npm install        # ajv, ajv-formats, @redocly/cli (node_modules is git-ignored; delete after use)
+node validate.mjs                        # 1-5 below
+npx redocly lint --config redocly.yaml ../../market/openapi/*.json   # 6
+node picker-test.mjs                     # 7 (deterministic half)
+node --test build.test.mjs               # builder unit tests
+```
+
+## 1-5. Format validation (`scripts/market/validate.mjs`)
+
+| # | Check | Standard used | Result |
+|---|---|---|---|
+| 1 | `catalog.json` against `schemas/catalog.schema.json` | JSON Schema 2020-12 (Ajv 8) | pass |
+| 2 | Every capability's input and output JSON Schema compiles; every example input validates against its input schema; every real output sample validates against its output schema | JSON Schema 2020-12 | pass (all README "Ready-made inputs" presets and `examples/` inputs are valid Actor inputs) |
+| 3 | `llms.txt` and `llms-full.txt` structure: H1, blockquote summary, H2 file lists of `- [name](url): notes` | [llmstxt.org](https://llmstxt.org/) v2 "Format" section | pass |
+| 4 | Every agent card (`.well-known/agent-card.json`, `agents/*.agent-card.json`) against `schemas/a2a-agent-card.schema.json` | Transcribed from the A2A Protocol Specification v1.0.0 §4.4-4.5 field tables (a2a-protocol.org, read 2026-09-24); discovery path `/.well-known/agent-card.json` per §8 | pass |
+| 5 | Every MCP manifest: `tools` validates as a `ListToolsResult` and each tool as a `Tool` | Official MCP JSON schema 2025-06-18 (`modelcontextprotocol/modelcontextprotocol` repo, `schema/2025-06-18/schema.json`) | pass |
+
+First run: 162 checks passed. Final run (with Mailvett, Bonafido, Fieldwright, web apps, the dataset API spec and the datasets
+MCP manifest): **189 checks passed** (52 capabilities, 11 agent cards, 26 MCP manifests).
+
+Honest limits: the A2A schema is our transcription of the spec tables (the spec's normative source is protobuf). Our cards
+declare MCP streamable HTTP and Apify's HTTP API as custom protocol bindings (spec §5.8 allows URIs); they are not native
+A2A JSON-RPC agents, and each card's description says so.
+
+## 6. OpenAPI lint
+
+Redocly CLI 2.x, `recommended` ruleset (only `info-license` off: the specs describe Apify's API, which has no per-product
+licence URL). Result: **all 11 specs valid, 0 errors, 0 warnings** (10 live Actors + `datasets.json`). The response code
+for synchronous runs (201) and the paths were checked against Apify's own OpenAPI 3.1 document
+(`https://docs.apify.com/api/openapi.json`). The hosted MCP tool names (`cybermax--<actor>`) were checked with a real
+`tools/list` call to `https://mcp.apify.com/?tools=cybermax/domain-intel` (no Actor run, no cost).
+
+## 7. Tool-picker test: can an agent choose the right capability from the descriptions?
+
+Twelve sample agent tasks in `tests/picker-tasks.json` (10 that map to a live tool, 1 that maps to a free dataset, 1 that
+nothing in the catalog should claim). Two independent pickers:
+
+**A. Blind LLM agent.** A separate Claude subagent was given only `market/llms.txt` (it could open `market/tools/*.md`),
+was told not to read the answer key, the scripts or the product folders, and had to name one catalog id (or `none`) per
+task with a confidence and a reason. Result: **12/12 correct**, confidence 0.85-0.98 (lowest: task 4, remote jobs search,
+0.85, because "HireHeat" and "HireHeat Jobs" look alike).
+
+**B. Deterministic BM25 baseline** (`scripts/market/picker-test.mjs`, re-runnable, exits 1 below 80%): ranks the live
+catalog by BM25 over brand, title, summary, agent description, use-when and keywords; a best score under 3.0 means "none".
+
+| Run | Result | Misses |
+|---|---|---|
+| First build | 9/12 (75%) | #4 remote jobs → HireHeat instead of HireHeat Jobs; #6 NVDA insider buys → the free Form 4 dataset instead of Boardroom Buys; #7 Duolingo complaints → the free complaint-stats dataset instead of Gripeline |
+| After description fixes | **10/12 (83%)** | #6 and #7: a keyword ranker can't read "not for live per-item lookups", so a same-topic snapshot dataset outscores the live tool. The LLM picker gets both right. |
+
+### Description fixes made because of the test (all legit, no keyword stuffing)
+
+- HireHeat vs HireHeat Jobs: `market.json` overrides give each a sharper summary, use-when and a "not for ... use the other one" line (the LLM picker's suggestion).
+- Every snapshot dataset whose brand matches a paid tool now says "not for live or per-item lookups: use the <brand> tool".
+- Summaries no longer cut off mid-word (Keyfern's did); Boardroom Buys' summary now names the ticker/lookback/open-market filters.
+- Dataset summaries now come from the card's first real paragraphs (they used to be a teaser question or a sentence ending in a colon), e.g. Pingroll, Kevscope, Open Domain Ranks.
+- Suggested but not applied here (Store titles belong to the growth lane): retitle HireHeat Jobs to "Job Search Across 12,000+ Career Sites" and HireHeat to "Company Hiring Momentum by Domain" so the two names stop colliding.
+
+### Final run (after adding web apps, Mailvett, Bonafido and Fieldwright to the catalog; 14 tasks)
+
+Two tasks were added: #13 "flag disposable and mistyped sign-up emails" (Mailvett, still coming soon) and #14 "a simple
+page to turn a salary into an hourly rate" (the Salary ↔ Hourly web app).
+
+- **Blind LLM agent: 14/14 correct** (confidence 0.75-0.95; #13 at 0.75 because Mailvett is marked "coming soon, not
+  callable yet" and the picker said so). It read only `llms.txt`.
+- **BM25: 11/14 (79%)** over 52 live + coming-soon entries. Misses: #6, #7 (dataset vs live tool, as before) and #12:
+  "translate this paragraph into French" now hits Hushscribe, whose description truthfully says "translate to English"
+  (Whisper audio translation). The LLM picker rejected it ("transcription, not translation"). The BM25 script is a
+  regression guard at the recorded 75% baseline; the LLM result is the bar.
+- Its remaining notes: the two "HireHeat" names (retitle suggestion above, growth lane); snapshot datasets vs live tools
+  (datasets now carry "not for live lookups"); datasets whose licence showed only "other" now show the licence name.
+
+Add a task to `tests/picker-tasks.json` whenever a new tool could be confused with an existing one, and re-run both pickers.

market/live-snapshot.json

@@ -1,5 +1,5 @@
 {
-  "fetchedAt": "2026-09-24T23:08:09.213Z",
+  "fetchedAt": "2026-09-24T23:25:23.259Z",
   "store": [
     {
       "name": "website-contacts",
@@ -532,11 +532,17 @@
   "datasets": [
     {
       "id": "shaw276/open-domain-ranks",
+      "splits": [
+        {
+          "config": "default",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 74,
       "prettyName": "Linkheft Open Domain Ranks (10.3M domains)",
       "description": "An open alternative to proprietary \"domain authority\" scores. For each of the top 10 million domains on the web (plus every Majestic Million domain) this dataset gives: domain, cc_rank, cc_pr_rank, cc_hosts, cc_prev_rank, mj_rank, mj_refsubnets, mj_refips.",
-      "license": "other",
+      "license": "other (cc-by-3.0-and-common-crawl-terms)",
       "tags": [
         "seo",
         "domain-authority",
@@ -553,6 +559,12 @@
     },
     {
       "id": "shaw276/saas-email-security-snapshot",
+      "splits": [
+        {
+          "config": "default",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 10,
       "prettyName": "DomainDNA: SaaS Email Security & Tech Stack Snapshot",
@@ -577,6 +589,12 @@
     },
     {
       "id": "shaw276/tech-hiring-snapshot",
+      "splits": [
+        {
+          "config": "default",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 15,
       "prettyName": "HireHeat: Tech Company Hiring Snapshot",
@@ -601,6 +619,12 @@
     },
     {
       "id": "shaw276/app-complaint-stats",
+      "splits": [
+        {
+          "config": "default",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 10,
       "prettyName": "Gripeline: App Store & Google Play Complaint Stats",
@@ -624,6 +648,12 @@
     },
     {
       "id": "shaw276/sec-form4-insider-transactions",
+      "splits": [
+        {
+          "config": "default",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 19,
       "prettyName": "Boardroom Buys: SEC Form 4 Insider Transactions (latest EDGAR filings)",
@@ -649,11 +679,25 @@
     },
     {
       "id": "shaw276/kevscope-cve-kev-epss-daily",
+      "splits": [
+        {
+          "config": "kev",
+          "split": "train"
+        },
+        {
+          "config": "epss",
+          "split": "train"
+        },
+        {
+          "config": "recent",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 21,
       "prettyName": "Kevscope: CVE + CISA KEV + EPSS, joined daily",
       "description": "Which vulnerabilities should you patch first? Kevscope puts the three public signals side by side, already joined on the CVE ID: CISA KEV: the US government's catalog of vulnerabilities known to be exploited in the wild, with due dates and ransomware use; EPSS (FIRST): the probability that a CVE will be exploited in the next 30 days, for every published CVE; NVD: CVSS severity (v4.0 → v3.1 → v3.0 → v2, best available), CWE weaknesses, affected vendor:product pairs and the description.",
-      "license": "other",
+      "license": "other (open-mixed-cc0-public-domain-epss-attribution)",
       "tags": [
         "cybersecurity",
         "vulnerabilities",
@@ -677,6 +721,16 @@
     },
     {
       "id": "shaw276/tickerbell-sec-8k-events",
+      "splits": [
+        {
+          "config": "events",
+          "split": "train"
+        },
+        {
+          "config": "filings",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 15,
       "prettyName": "Tickerbell: SEC 8-K Material Events by Item (latest EDGAR filings)",
@@ -704,6 +758,12 @@
     },
     {
       "id": "shaw276/us-zip-codes-demographics",
+      "splits": [
+        {
+          "config": "default",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 21,
       "prettyName": "Ziplore US ZIP Codes with Census Demographics (41,682 ZIPs)",
@@ -726,6 +786,20 @@
     },
     {
       "id": "shaw276/curvewire-us-macro-panel",
+      "splits": [
+        {
+          "config": "treasury_yields",
+          "split": "train"
+        },
+        {
+          "config": "bls_monthly",
+          "split": "train"
+        },
+        {
+          "config": "treasury_real_yields",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 12,
       "prettyName": "Curvewire: US Treasury yield curve + BLS macro panel (official sources)",
@@ -754,11 +828,17 @@
     },
     {
       "id": "shaw276/trialtide-clinicaltrials-registry",
+      "splits": [
+        {
+          "config": "default",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 10,
       "prettyName": "Trialtide: every ClinicalTrials.gov study in one table",
       "description": "All 604,146 studies registered on ClinicalTrials.gov, flattened to one row per study: status, phase, design, enrollment, key dates, conditions (with MeSH terms), interventions, lead sponsor and sponsor type, collaborators, countries, number of sites, eligibility basics, primary outcomes and whether results are posted.",
-      "license": "other",
+      "license": "other (us-public-domain-clinicaltrials-gov)",
       "tags": [
         "clinical-trials",
         "clinicaltrials-gov",
@@ -778,6 +858,52 @@
     },
     {
       "id": "shaw276/fipsbook-us-census-geographies",
+      "splits": [
+        {
+          "config": "counties",
+          "split": "train"
+        },
+        {
+          "config": "congressional_districts",
+          "split": "train"
+        },
+        {
+          "config": "county_subdivisions",
+          "split": "train"
+        },
+        {
+          "config": "metro_areas",
+          "split": "train"
+        },
+        {
+          "config": "places",
+          "split": "train"
+        },
+        {
+          "config": "school_districts",
+          "split": "train"
+        },
+        {
+          "config": "state_house_districts",
+          "split": "train"
+        },
+        {
+          "config": "state_senate_districts",
+          "split": "train"
+        },
+        {
+          "config": "states",
+          "split": "train"
+        },
+        {
+          "config": "tracts",
+          "split": "train"
+        },
+        {
+          "config": "zctas",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 0,
       "prettyName": "Fipsbook: US Census Geographies with FIPS Codes, Centroids & Land Area (2025)",
@@ -814,11 +940,17 @@
     },
     {
       "id": "shaw276/toolroll-mcp-registry-snapshot",
+      "splits": [
+        {
+          "config": "servers",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 0,
       "prettyName": "Toolroll: Official MCP Registry Snapshot (every MCP server, weekly)",
       "description": "A snapshot of the official Model Context Protocol (MCP) Registry (registry.modelcontextprotocol.io) as a single Parquet table: 35,403 MCP servers (35,028 active, 375 deprecated) from 20,000+ publishers, snapshot 2026-09-24 07:03 UTC. One row per server (its latest version) with its description, how you connect to it (remote URL and transport, or npm / PyPI / Docker (OCI) / MCPB / NuGet / Cargo package), whether it needs a secret (API key header or secret env var), repository, website and publish dates.",
-      "license": "other",
+      "license": "other (mcp-registry-public-metadata)",
       "tags": [
         "mcp",
         "model-context-protocol",
@@ -838,6 +970,20 @@
     },
     {
       "id": "shaw276/tallyfold-sec-xbrl-financials",
+      "splits": [
+        {
+          "config": "key_metrics",
+          "split": "train"
+        },
+        {
+          "config": "facts",
+          "split": "train"
+        },
+        {
+          "config": "filings",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 0,
       "prettyName": "Tallyfold: SEC XBRL Financial Statements & Key Metrics (every US filer, latest 4 quarters)",
@@ -866,6 +1012,12 @@
     },
     {
       "id": "shaw276/pingroll-mcp-server-uptime",
+      "splits": [
+        {
+          "config": "default",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 0,
       "prettyName": "Pingroll: MCP Server Uptime (daily handshake probes of every remote server in the official MCP Registry)",
@@ -891,6 +1043,16 @@
     },
     {
       "id": "shaw276/riskroll-sec-10k-10q-sections",
+      "splits": [
+        {
+          "config": "sections",
+          "split": "train"
+        },
+        {
+          "config": "filings",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 0,
       "prettyName": "Riskroll: SEC 10-K & 10-Q Risk Factors, MD&A and Market Risk Text (clean, by section)",
@@ -918,6 +1080,24 @@
     },
     {
       "id": "shaw276/policyloom-us-ai-policy-tracker",
+      "splits": [
+        {
+          "config": "documents",
+          "split": "train"
+        },
+        {
+          "config": "open_comment_periods",
+          "split": "train"
+        },
+        {
+          "config": "public_inspection",
+          "split": "train"
+        },
+        {
+          "config": "weekly_counts",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 0,
       "prettyName": "Policyloom: US AI Policy Tracker (Federal Register, daily)",
@@ -947,11 +1127,17 @@
     },
     {
       "id": "shaw276/swellmeter-google-trending-searches",
+      "splits": [
+        {
+          "config": "default",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 0,
       "prettyName": "Swellmeter: Google Trending Searches (daily snapshot, 30 countries)",
       "description": "A daily snapshot of Google Trends' \"Trending now\" list for 30 countries: each trending search, its approximate traffic bucket, when it started trending, and the news articles Google attached to it. One Parquet file per day (data/trending-YYYY-MM-DD.parquet), so a history of what the world searched for builds up over time. Google's own feed only shows what is trending now; this dataset keeps it.",
-      "license": "other",
+      "license": "other (google-trends-public-feed)",
       "tags": [
         "google-trends",
         "trending-searches",
@@ -970,11 +1156,17 @@
     },
     {
       "id": "shaw276/hireheat-career-site-jobs-weekly",
+      "splits": [
+        {
+          "config": "default",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 0,
       "prettyName": "HireHeat Jobs: career-site job postings, first posted in the last 7 days (weekly snapshots)",
       "description": "Each snapshot holds every job first posted in the last 7 days on the company career sites in the HireHeat Jobs directory, read straight from each employer's applicant tracking system (Greenhouse, Ashby, Workable, SmartRecruiters). One row per job with title, company, location, remote flag, department, employment type, salary where the employer publishes it, posting date and the apply link, plus a derived job function and seniority. No description text and no people: job postings only. A new file is added every week, so the history of who hires what builds up.",
-      "license": "other",
+      "license": "other (public-job-postings)",
       "tags": [
         "jobs",
         "job-postings",
@@ -996,6 +1188,12 @@
     },
     {
       "id": "shaw276/smartthings-review-mood-sept-2026",
+      "splits": [
+        {
+          "config": "default",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 0,
       "prettyName": "Gripeline: SmartThings App Review Mood Around the Sept 2026 Fridge Firmware Incident",
@@ -1020,6 +1218,12 @@
     },
     {
       "id": "shaw276/superintelligence-search-questions",
+      "splits": [
+        {
+          "config": "default",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 0,
       "prettyName": "Keyfern: What People Search About Superintelligence (Sept 2026)",
@@ -1043,6 +1247,20 @@
     },
     {
       "id": "shaw276/superintelligence-ban-bill-coverage",
+      "splits": [
+        {
+          "config": "headlines",
+          "split": "train"
+        },
+        {
+          "config": "daily",
+          "split": "train"
+        },
+        {
+          "config": "sources",
+          "split": "train"
+        }
+      ],
       "likes": 0,
       "downloads": 0,
       "prettyName": "Policyloom: Superintelligence Ban Bill, News Coverage & Search Interest (Sept 2026)",
@@ -1081,5 +1299,67 @@
     "io.github.shawmena-dotcom/sec-insider-tracker",
     "io.github.shawmena-dotcom/website-contacts",
     "io.github.shawmena-dotcom/website-screenshot"
+  ],
+  "webApps": [
+    {
+      "id": "sumbloom",
+      "name": "Sumbloom",
+      "factory": "Games",
+      "category": "Daily number puzzle game",
+      "url": "https://shaw276-cybermax.static.hf.space/sumbloom/index.html",
+      "title": "Sumbloom — Free Daily Number Puzzle (Math Game Like Digits)",
+      "description": "Free daily number puzzle: combine six numbers with + − × ÷ to hit five targets. New puzzle every day, shareable results, streaks and practice.",
+      "space": "https://huggingface.co/spaces/shaw276/sumbloom"
+    },
+    {
+      "id": "honeyfold",
+      "name": "Honeyfold",
+      "factory": "Games",
+      "category": "Idle / incremental game",
+      "url": "https://shaw276-cybermax.static.hf.space/honeyfold/index.html",
+      "title": "Honeyfold — Free Idle Bee Game You Can Finish",
+      "description": "Honeyfold is a free idle clicker game: grow a bee colony from one bee to the Golden Comb in about 30 minutes. Offline progress, golden flowers and New Game+.",
+      "space": null
+    },
+    {
+      "id": "utm-builder",
+      "name": "UTM Builder",
+      "factory": "Apps",
+      "category": "Marketing web app",
+      "url": "https://shaw276-cybermax.static.hf.space/utm-builder/index.html",
+      "title": "Free UTM Builder — Bulk Campaign URL Builder + QR Code",
+      "description": "Free UTM builder: tag campaign URLs for Google Analytics 4, build hundreds at once from a list or CSV, make a QR code, and decode tagged links. No sign-up.",
+      "space": null
+    },
+    {
+      "id": "salary-to-hourly-calculator",
+      "name": "Salary ↔ Hourly Calculator",
+      "factory": "Apps",
+      "category": "Finance calculator",
+      "url": "https://shaw276-cybermax.static.hf.space/salary-to-hourly-calculator/index.html",
+      "title": "Salary to Hourly Calculator + Freelance Rate Calculator",
+      "description": "Free salary to hourly calculator: convert annual pay to hourly, daily, weekly and monthly, hourly to salary with overtime, and work out your freelance rate.",
+      "space": null
+    },
+    {
+      "id": "profit-margin-calculator",
+      "name": "Profit Margin Calculator",
+      "factory": "Apps",
+      "category": "Business calculator",
+      "url": "https://shaw276-cybermax.static.hf.space/profit-margin-calculator/index.html",
+      "title": "Profit Margin Calculator + Markup Calculator (Free)",
+      "description": "Free profit margin calculator: enter any two of cost, price, profit, margin or markup and get the rest. Net margin after fees, discounts and a product table.",
+      "space": null
+    },
+    {
+      "id": "hushscribe",
+      "name": "Hushscribe",
+      "factory": "AI",
+      "category": "Private in-browser transcription (Whisper)",
+      "url": "https://shaw276-hushscribe.static.hf.space/index.html",
+      "title": "Hushscribe — Private Audio Transcription",
+      "description": "Free private audio and video transcription that runs in your browser with Whisper. No upload, no sign-up. Export TXT, SRT subtitles, VTT and JSON. 99 languages, translate to English. By CyberMax.",
+      "space": "https://huggingface.co/spaces/shaw276/hushscribe"
+    }
   ]
 }

market/llms-full.txt

@@ -2,7 +2,7 @@
 
 > Machine-readable catalog of every live CyberMax capability: 10 pay-per-result API tools (callable as MCP tools or HTTP APIs, billed to the caller's own Apify account) and 21 free open datasets. Pick a tool by what it returns; every entry has JSON Schemas for input and output, pricing and copy-paste integration code.
 
-How to call a tool: hosted MCP `https://mcp.apify.com/?tools=cybermax/<name>` (streamable HTTP, header `Authorization: Bearer <APIFY_TOKEN>`; several tools: comma-separate them), tool name `cybermax--<name>`; or HTTP `POST https://api.apify.com/v2/acts/cybermax~<name>/run-sync-get-dataset-items` with the input JSON. Structured catalog: https://shaw276-cybermax.static.hf.space/market/catalog.json. Built 2026-09-24T23:08:09.213Z.
+How to call a tool: hosted MCP `https://mcp.apify.com/?tools=cybermax/<name>` (streamable HTTP, header `Authorization: Bearer <APIFY_TOKEN>`; several tools: comma-separate them), tool name `cybermax--<name>`; or HTTP `POST https://api.apify.com/v2/acts/cybermax~<name>/run-sync-get-dataset-items` with the input JSON. Structured catalog: https://shaw276-cybermax.static.hf.space/market/catalog.json. Built 2026-09-24T23:25:23.259Z.
 
 ## Tools
 - [Gripeline](https://apify.com/cybermax/app-reviews): App Store and Google Play reviews scraper: both stores in one clean schema, plus a per-app summary agents can read (star mix, % negative, top complaints, worst app versions).
@@ -17,24 +17,24 @@ How to call a tool: hosted MCP `https://mcp.apify.com/?tools=cybermax/<name>` (s
 - [SnapScout](https://apify.com/cybermax/website-screenshot): Website screenshot API for AI agents: one record per URL with a public PNG/JPEG screenshot (full page, element or mobile), HTTP status, title, meta and clean page text.
 
 ## Datasets
-- [Linkheft Open Domain Ranks (10.3M domains)](https://huggingface.co/datasets/shaw276/open-domain-ranks): An open alternative to proprietary "domain authority" scores. For each of the top 10 million domains on the web (plus every Majestic Million domain) this dataset gives: domain, cc_rank, cc_pr_rank, cc_hosts, cc_prev_rank, mj_rank, mj_refsubnets, mj_refips. License: other.
+- [Linkheft Open Domain Ranks (10.3M domains)](https://huggingface.co/datasets/shaw276/open-domain-ranks): An open alternative to proprietary "domain authority" scores. For each of the top 10 million domains on the web (plus every Majestic Million domain) this dataset gives: domain, cc_rank, cc_pr_rank, cc_hosts, cc_prev_rank, mj_rank, mj_refsubnets, mj_refips. License: other (cc-by-3.0-and-common-crawl-terms).
 - [DomainDNA: SaaS Email Security & Tech Stack Snapshot](https://huggingface.co/datasets/shaw276/saas-email-security-snapshot): How do well-known SaaS companies protect their email domains and what do their websites run on? License: cc-by-4.0.
 - [HireHeat: Tech Company Hiring Snapshot](https://huggingface.co/datasets/shaw276/tech-hiring-snapshot): Who is hiring in tech right now, and for what? One row per company per snapshot with open-job counts from each company's own public job board (Greenhouse, Lever, Ashby, SmartRecruiters, Workable, Recruitee), broken down by function, seniority, top departments and top locations, plus how many roles… License: cc-by-4.0.
 - [Gripeline: App Store & Google Play Complaint Stats](https://huggingface.co/datasets/shaw276/app-complaint-stats): What are users of popular apps complaining about right now? One row per app per store per snapshot, computed from the newest reviews (up to 300 per app and store, US App Store and English Google Play): star mix (1 to 5), average rating of the sample, share of negative (1-2 stars) and positive (4-5… License: cc-by-4.0.
 - [Boardroom Buys: SEC Form 4 Insider Transactions (latest EDGAR filings)](https://huggingface.co/datasets/shaw276/sec-form4-insider-transactions): Every insider stock transaction (non-derivative table) reported to the SEC on Form 4 and 4/A over the latest 5 EDGAR business days (filed 2026-09-17 to 2026-09-23; a few late filings report older trades): officers, directors and 10% owners buying, selling, receiving grants, exercising options… License: cc0-1.0.
-- [Kevscope: CVE + CISA KEV + EPSS, joined daily](https://huggingface.co/datasets/shaw276/kevscope-cve-kev-epss-daily): Which vulnerabilities should you patch first? Kevscope puts the three public signals side by side, already joined on the CVE ID: CISA KEV: the US government's catalog of vulnerabilities known to be exploited in the wild, with due dates and ransomware use; EPSS (FIRST): the probability that a CVE… License: other.
+- [Kevscope: CVE + CISA KEV + EPSS, joined daily](https://huggingface.co/datasets/shaw276/kevscope-cve-kev-epss-daily): Which vulnerabilities should you patch first? Kevscope puts the three public signals side by side, already joined on the CVE ID: CISA KEV: the US government's catalog of vulnerabilities known to be exploited in the wild, with due dates and ransomware use; EPSS (FIRST): the probability that a CVE… License: other (open-mixed-cc0-public-domain-epss-attribution).
 - [Tickerbell: SEC 8-K Material Events by Item (latest EDGAR filings)](https://huggingface.co/datasets/shaw276/tickerbell-sec-8k-events): Every current report (Form 8-K and 8-K/A) filed with the SEC over the latest 20 EDGAR business days, split into one row per reported Item: new material agreements (1.01), earnings releases (2.02), acquisitions closed (2.01), impairments (2.06), delisting notices (3.01), auditor changes (4.01)… License: cc0-1.0.
 - [Ziplore US ZIP Codes with Census Demographics (41,682 ZIPs)](https://huggingface.co/datasets/shaw276/us-zip-codes-demographics): A free, current replacement for the frozen ZIP data inside uszipcode (last release 2022), pyzipcode (2021) and the zipcodes npm package (2018). License: cc-by-4.0.
 - [Curvewire: US Treasury yield curve + BLS macro panel (official sources)](https://huggingface.co/datasets/shaw276/curvewire-us-macro-panel): The US numbers macro models, trading bots and LLM agents ask for most, taken straight from the agencies that publish them (US Treasury and the Bureau of Labor Statistics), not scraped from redistributors, and cleaned into three tidy tables: treasury_yields (default), treasury_real_yields… License: cc0-1.0.
-- [Trialtide: every ClinicalTrials.gov study in one table](https://huggingface.co/datasets/shaw276/trialtide-clinicaltrials-registry): All 604,146 studies registered on ClinicalTrials.gov, flattened to one row per study: status, phase, design, enrollment, key dates, conditions (with MeSH terms), interventions, lead sponsor and sponsor type, collaborators, countries, number of sites, eligibility basics, primary outcomes and whether… License: other.
+- [Trialtide: every ClinicalTrials.gov study in one table](https://huggingface.co/datasets/shaw276/trialtide-clinicaltrials-registry): All 604,146 studies registered on ClinicalTrials.gov, flattened to one row per study: status, phase, design, enrollment, key dates, conditions (with MeSH terms), interventions, lead sponsor and sponsor type, collaborators, countries, number of sites, eligibility basics, primary outcomes and whether… License: other (us-public-domain-clinicaltrials-gov).
 - [Fipsbook: US Census Geographies with FIPS Codes, Centroids & Land Area (2025)](https://huggingface.co/datasets/shaw276/fipsbook-us-census-geographies): Free lookup tables for joining, mapping and geocoding US data: 210,319 rows across 11 layers (states, 3,222 counties, 32,350 places, 36,427 county subdivisions, 33,791 ZIP code areas (ZCTAs), 85,396 Census tracts, 935 metro/micro areas, 440 congressional districts of the 119th Congress, 6,843 state… License: cc0-1.0.
-- [Toolroll: Official MCP Registry Snapshot (every MCP server, weekly)](https://huggingface.co/datasets/shaw276/toolroll-mcp-registry-snapshot): A snapshot of the official Model Context Protocol (MCP) Registry (registry.modelcontextprotocol.io) as a single Parquet table: 35,403 MCP servers (35,028 active, 375 deprecated) from 20,000+ publishers, snapshot 2026-09-24 07:03 UTC. License: other.
+- [Toolroll: Official MCP Registry Snapshot (every MCP server, weekly)](https://huggingface.co/datasets/shaw276/toolroll-mcp-registry-snapshot): A snapshot of the official Model Context Protocol (MCP) Registry (registry.modelcontextprotocol.io) as a single Parquet table: 35,403 MCP servers (35,028 active, 375 deprecated) from 20,000+ publishers, snapshot 2026-09-24 07:03 UTC. License: other (mcp-registry-public-metadata).
 - [Tallyfold: SEC XBRL Financial Statements & Key Metrics (every US filer, latest 4 quarters)](https://huggingface.co/datasets/shaw276/tallyfold-sec-xbrl-financials): The numbers from every financial statement filed with the SEC in the latest four quarters: 10-K, 10-Q, 20-F and 40-F reports from 7,000+ US-listed and SEC-registered companies, as tidy parquet you can query in one line. License: cc0-1.0.
 - [Pingroll: MCP Server Uptime (daily handshake probes of every remote server in the official MCP Registry)](https://huggingface.co/datasets/shaw276/pingroll-mcp-server-uptime): Registries and directories tell you an MCP server exists. This dataset tells you whether it works: every public remote server in the official MCP Registry gets a real MCP handshake once a day (initialize → notifications/initialized → tools/list; no tool is ever called), and each result is recorded… License: cc0-1.0.
 - [Riskroll: SEC 10-K & 10-Q Risk Factors, MD&A and Market Risk Text (clean, by section)](https://huggingface.co/datasets/shaw276/riskroll-sec-10k-10q-sections): The parts of annual and quarterly reports that analysts, researchers and LLM pipelines actually read, cut out of each filing and cleaned: Risk Factors (Item 1A), Management's Discussion and Analysis (MD&A), Quantitative and Qualitative Disclosures About Market Risk, Cybersecurity (Item 1C) and… License: cc0-1.0.
 - [Policyloom: US AI Policy Tracker (Federal Register, daily)](https://huggingface.co/datasets/shaw276/policyloom-us-ai-policy-tracker): Every document in the US Federal Register since 2017-01-01 whose text mentions artificial intelligence: executive orders and other presidential documents, final rules, proposed rules, notices, requests for information and their comment periods, with dates, agencies, dockets and links. License: cc0-1.0.
-- [Swellmeter: Google Trending Searches (daily snapshot, 30 countries)](https://huggingface.co/datasets/shaw276/swellmeter-google-trending-searches): A daily snapshot of Google Trends' "Trending now" list for 30 countries: each trending search, its approximate traffic bucket, when it started trending, and the news articles Google attached to it. License: other.
-- [HireHeat Jobs: career-site job postings, first posted in the last 7 days (weekly snapshots)](https://huggingface.co/datasets/shaw276/hireheat-career-site-jobs-weekly): Each snapshot holds every job first posted in the last 7 days on the company career sites in the HireHeat Jobs directory, read straight from each employer's applicant tracking system (Greenhouse, Ashby, Workable, SmartRecruiters). License: other.
+- [Swellmeter: Google Trending Searches (daily snapshot, 30 countries)](https://huggingface.co/datasets/shaw276/swellmeter-google-trending-searches): A daily snapshot of Google Trends' "Trending now" list for 30 countries: each trending search, its approximate traffic bucket, when it started trending, and the news articles Google attached to it. License: other (google-trends-public-feed).
+- [HireHeat Jobs: career-site job postings, first posted in the last 7 days (weekly snapshots)](https://huggingface.co/datasets/shaw276/hireheat-career-site-jobs-weekly): Each snapshot holds every job first posted in the last 7 days on the company career sites in the HireHeat Jobs directory, read straight from each employer's applicant tracking system (Greenhouse, Ashby, Workable, SmartRecruiters). License: other (public-job-postings).
 - [Gripeline: SmartThings App Review Mood Around the Sept 2026 Fridge Firmware Incident](https://huggingface.co/datasets/shaw276/smartthings-review-mood-sept-2026): On 22 Sep 2026 a refrigerator firmware update pushed through Samsung's SmartThings platform left some smart fridges in Korea without power (Ars Technica, 23 Sep 2026; Samsung told Ars the issue was limited to Korea and came from an error during internal testing). License: cc-by-4.0.
 - [Keyfern: What People Search About Superintelligence (Sept 2026)](https://huggingface.co/datasets/shaw276/superintelligence-search-questions): "Superintelligence" moved from books and sci-fi into the news in the week of 22 Sep 2026: an AI remark in a UN General Assembly speech (22 Sep) and a proposed US bill to ban artificial superintelligence (reported by AP on 23 Sep; see the Policyloom AI policy tracker for the verified events). License: cc-by-4.0.
 - [Policyloom: Superintelligence Ban Bill, News Coverage & Search Interest (Sept 2026)](https://huggingface.co/datasets/shaw276/superintelligence-ban-bill-coverage): In September 2026 "superintelligence" became a US policy topic: a US senator and a House member announced a bill to ban the development of artificial superintelligence and pause advanced AI development until a federal regulator sets safety rules (Senate press release, 3 Sep; widely reported on 23… License: cc-by-4.0.
@@ -267,14 +267,26 @@ Example input:
 ```
 
 
+## Web apps
+
+- [Sumbloom — Free Daily Number Puzzle (Math Game Like Digits)](https://shaw276-cybermax.static.hf.space/sumbloom/index.html): Free daily number puzzle: combine six numbers with + − × ÷ to hit five targets. Free, for people (no API).
+- [Honeyfold — Free Idle Bee Game You Can Finish](https://shaw276-cybermax.static.hf.space/honeyfold/index.html): Honeyfold is a free idle clicker game: grow a bee colony from one bee to the Golden Comb in about 30 minutes. Free, for people (no API).
+- [Free UTM Builder — Bulk Campaign URL Builder + QR Code](https://shaw276-cybermax.static.hf.space/utm-builder/index.html): Free UTM builder: tag campaign URLs for Google Analytics 4, build hundreds at once from a list or CSV, make a QR code, and decode tagged links. Free, for people (no API).
+- [Salary to Hourly Calculator + Freelance Rate Calculator](https://shaw276-cybermax.static.hf.space/salary-to-hourly-calculator/index.html): Free salary to hourly calculator: convert annual pay to hourly, daily, weekly and monthly, hourly to salary with overtime, and work out your freelance rate. Free, for people (no API).
+- [Profit Margin Calculator + Markup Calculator (Free)](https://shaw276-cybermax.static.hf.space/profit-margin-calculator/index.html): Free profit margin calculator: enter any two of cost, price, profit, margin or markup and get the rest. Free, for people (no API).
+- [Hushscribe — Private Audio Transcription](https://shaw276-hushscribe.static.hf.space/index.html): Free private audio and video transcription that runs in your browser with Whisper. Free, for people (no API).
+
 ## Optional
 
+- [Fieldwright (coming soon, not callable yet): AI Web Data Extractor (Bring Your Own LLM Key)](https://shaw276-cybermax.static.hf.space/market/tools/ai-extract.md): AI web scraper that turns any page into typed JSON fields you name, using YOUR OpenAI, Claude, Gemini or OpenRouter key (you pay the model at cost).
 - [Inkpluck (coming soon, not callable yet): Article Extractor: URL to Text, Markdown & Metadata](https://shaw276-cybermax.static.hf.space/market/tools/article-extract.md): Article extractor API for news, blogs and docs: URLs or HTML in, clean text + Markdown, title, authors, publish date, language, top image, videos and paywall flag out.
 - [Pinloft (coming soon, not callable yet): Batch Geocoder: Address to Lat/Long + Census FIPS](https://shaw276-cybermax.static.hf.space/market/tools/batch-geocoder.md): Batch geocoding API: addresses to latitude/longitude, cleaned address, county, Census tract and block FIPS for US addresses (US Census), worldwide via OpenStreetMap.
 - [Picpail (coming soon, not callable yet): Bulk Image Downloader: All Images from URLs to ZIP](https://shaw276-cybermax.static.hf.space/market/tools/bulk-image-downloader.md): Download all images from web pages or image URLs into one ZIP, full-size originals instead of thumbnails: img, srcset, lazy-load, CSS, og:image.
 - [Firmfold (coming soon, not callable yet): Company Enrichment API: Logo, Tech, Contacts, Jobs](https://shaw276-cybermax.static.hf.space/market/tools/company-enrichment.md): Company enrichment API: give domains, work emails or company names; get one flat profile each: official domain, logo + brand colour, tech stack, email security, domain age, contacts, socials, hiring and Wikidata firmographics.
 - [Namewhere (coming soon, not callable yet): Company Name to Domain & Website Finder](https://shaw276-cybermax.static.hf.space/market/tools/company-to-domain.md): Company name to domain finder: turn company names into official website domains with a confidence score and evidence (Wikidata, homepage brand check), alternatives, social links, parked domains rejected.
 - [Commonkite (coming soon, not callable yet): Free Image Search API: Creative Commons & CC0](https://shaw276-cybermax.static.hf.space/market/tools/creative-commons-images.md): Search free-to-use images by keyword across Openverse (900M+ CC images), NASA, The Met, Art Institute of Chicago, Cleveland Museum of Art and Wikimedia Commons.
+- [Bonafido (coming soon, not callable yet): Business Domain Verification API](https://shaw276-cybermax.static.hf.space/market/tools/domain-verify.md): Checks whether a domain, URL or email domain is a real, operating, legitimate business website and (optionally) whether it belongs to the company you expect; returns a verdict, a 0-100 trust score, machine-readable risk flags with reasons, and quotable evidence.
+- [Mailvett (coming soon, not callable yet): Email Validation API for AI Agents](https://shaw276-cybermax.static.hf.space/market/tools/email-validate.md): Validate email addresses before sending or saving them: syntax, domain and MX / mail provider, disposable, role, free-provider and typo flags, domain age, SPF/DMARC and parked-domain signals, returned as a verdict (deliverable_likely, risky, undeliverable, invalid) with reason codes and a 0-100…
 - [Printwren (coming soon, not callable yet): HTML to PDF & URL to PDF API](https://shaw276-cybermax.static.hf.space/market/tools/html-to-pdf.md): HTML to PDF and URL to PDF API with headless Chrome: A4/Letter/Legal, landscape, margins, header/footer with page numbers, print CSS.
 - [Logolark (coming soon, not callable yet): Company Logo API & Favicon Finder by Domain](https://shaw276-cybermax.static.hf.space/market/tools/logo-finder.md): Clearbit Logo API replacement: give domains, URLs or emails, get each company's own best logo and square icon, verified by download: real size, SVG/PNG/ICO, transparency, brand colour, hosted copy, plus rebrand alerts in monitor mode.
 - [Paperkiln (coming soon, not callable yet): Word, Excel & PowerPoint to PDF Converter API](https://shaw276-cybermax.static.hf.space/market/tools/office-to-pdf.md): Convert DOCX, XLSX, PPTX, ODT, RTF, CSV and legacy DOC/XLS/PPT files to PDF (or PDF/A) in bulk with headless LibreOffice and Office-compatible fonts.

market/llms.txt

@@ -2,7 +2,7 @@
 
 > Machine-readable catalog of every live CyberMax capability: 10 pay-per-result API tools (callable as MCP tools or HTTP APIs, billed to the caller's own Apify account) and 21 free open datasets. Pick a tool by what it returns; every entry has JSON Schemas for input and output, pricing and copy-paste integration code.
 
-How to call a tool: hosted MCP `https://mcp.apify.com/?tools=cybermax/<name>` (streamable HTTP, header `Authorization: Bearer <APIFY_TOKEN>`; several tools: comma-separate them), tool name `cybermax--<name>`; or HTTP `POST https://api.apify.com/v2/acts/cybermax~<name>/run-sync-get-dataset-items` with the input JSON. Structured catalog: https://shaw276-cybermax.static.hf.space/market/catalog.json. Built 2026-09-24T23:08:09.213Z.
+How to call a tool: hosted MCP `https://mcp.apify.com/?tools=cybermax/<name>` (streamable HTTP, header `Authorization: Bearer <APIFY_TOKEN>`; several tools: comma-separate them), tool name `cybermax--<name>`; or HTTP `POST https://api.apify.com/v2/acts/cybermax~<name>/run-sync-get-dataset-items` with the input JSON. Structured catalog: https://shaw276-cybermax.static.hf.space/market/catalog.json. Built 2026-09-24T23:25:23.259Z.
 
 ## Tools
 - [Gripeline: App Store & Google Play Reviews Scraper](https://shaw276-cybermax.static.hf.space/market/tools/app-reviews.md): App Store and Google Play reviews scraper: both stores in one clean schema, plus a per-app summary agents can read (star mix, % negative, top complaints, worst app versions). Price: $0.003 per app summary + $0.0001 per review. MCP tool `cybermax--app-reviews`.
@@ -17,36 +17,48 @@ How to call a tool: hosted MCP `https://mcp.apify.com/?tools=cybermax/<name>` (s
 - [SnapScout: Website Screenshot API for AI Agents (Image + Text)](https://shaw276-cybermax.static.hf.space/market/tools/website-screenshot.md): Website screenshot API for AI agents: one record per URL with a public PNG/JPEG screenshot (full page, element or mobile), HTTP status, title, meta and clean page text. Price: $0.003 per screenshot captured. MCP tool `cybermax--website-screenshot`.
 
 ## Datasets
-- [Linkheft Open Domain Ranks (10.3M domains)](https://huggingface.co/datasets/shaw276/open-domain-ranks): An open alternative to proprietary "domain authority" scores. For each of the top 10 million domains on the web (plus every Majestic Million domain) this dataset gives: domain, cc_rank, cc_pr_rank, cc_hosts, cc_prev_rank, mj_rank, mj_refsubnets, mj_refips. License: other.
+- [Linkheft Open Domain Ranks (10.3M domains)](https://huggingface.co/datasets/shaw276/open-domain-ranks): An open alternative to proprietary "domain authority" scores. For each of the top 10 million domains on the web (plus every Majestic Million domain) this dataset gives: domain, cc_rank, cc_pr_rank, cc_hosts, cc_prev_rank, mj_rank, mj_refsubnets, mj_refips. License: other (cc-by-3.0-and-common-crawl-terms).
 - [DomainDNA: SaaS Email Security & Tech Stack Snapshot](https://huggingface.co/datasets/shaw276/saas-email-security-snapshot): How do well-known SaaS companies protect their email domains and what do their websites run on? License: cc-by-4.0.
 - [HireHeat: Tech Company Hiring Snapshot](https://huggingface.co/datasets/shaw276/tech-hiring-snapshot): Who is hiring in tech right now, and for what? One row per company per snapshot with open-job counts from each company's own public job board (Greenhouse, Lever, Ashby, SmartRecruiters, Workable, Recruitee), broken down by function, seniority, top departments and top locations, plus how many roles… License: cc-by-4.0.
 - [Gripeline: App Store & Google Play Complaint Stats](https://huggingface.co/datasets/shaw276/app-complaint-stats): What are users of popular apps complaining about right now? One row per app per store per snapshot, computed from the newest reviews (up to 300 per app and store, US App Store and English Google Play): star mix (1 to 5), average rating of the sample, share of negative (1-2 stars) and positive (4-5… License: cc-by-4.0.
 - [Boardroom Buys: SEC Form 4 Insider Transactions (latest EDGAR filings)](https://huggingface.co/datasets/shaw276/sec-form4-insider-transactions): Every insider stock transaction (non-derivative table) reported to the SEC on Form 4 and 4/A over the latest 5 EDGAR business days (filed 2026-09-17 to 2026-09-23; a few late filings report older trades): officers, directors and 10% owners buying, selling, receiving grants, exercising options… License: cc0-1.0.
-- [Kevscope: CVE + CISA KEV + EPSS, joined daily](https://huggingface.co/datasets/shaw276/kevscope-cve-kev-epss-daily): Which vulnerabilities should you patch first? Kevscope puts the three public signals side by side, already joined on the CVE ID: CISA KEV: the US government's catalog of vulnerabilities known to be exploited in the wild, with due dates and ransomware use; EPSS (FIRST): the probability that a CVE… License: other.
+- [Kevscope: CVE + CISA KEV + EPSS, joined daily](https://huggingface.co/datasets/shaw276/kevscope-cve-kev-epss-daily): Which vulnerabilities should you patch first? Kevscope puts the three public signals side by side, already joined on the CVE ID: CISA KEV: the US government's catalog of vulnerabilities known to be exploited in the wild, with due dates and ransomware use; EPSS (FIRST): the probability that a CVE… License: other (open-mixed-cc0-public-domain-epss-attribution).
 - [Tickerbell: SEC 8-K Material Events by Item (latest EDGAR filings)](https://huggingface.co/datasets/shaw276/tickerbell-sec-8k-events): Every current report (Form 8-K and 8-K/A) filed with the SEC over the latest 20 EDGAR business days, split into one row per reported Item: new material agreements (1.01), earnings releases (2.02), acquisitions closed (2.01), impairments (2.06), delisting notices (3.01), auditor changes (4.01)… License: cc0-1.0.
 - [Ziplore US ZIP Codes with Census Demographics (41,682 ZIPs)](https://huggingface.co/datasets/shaw276/us-zip-codes-demographics): A free, current replacement for the frozen ZIP data inside uszipcode (last release 2022), pyzipcode (2021) and the zipcodes npm package (2018). License: cc-by-4.0.
 - [Curvewire: US Treasury yield curve + BLS macro panel (official sources)](https://huggingface.co/datasets/shaw276/curvewire-us-macro-panel): The US numbers macro models, trading bots and LLM agents ask for most, taken straight from the agencies that publish them (US Treasury and the Bureau of Labor Statistics), not scraped from redistributors, and cleaned into three tidy tables: treasury_yields (default), treasury_real_yields… License: cc0-1.0.
-- [Trialtide: every ClinicalTrials.gov study in one table](https://huggingface.co/datasets/shaw276/trialtide-clinicaltrials-registry): All 604,146 studies registered on ClinicalTrials.gov, flattened to one row per study: status, phase, design, enrollment, key dates, conditions (with MeSH terms), interventions, lead sponsor and sponsor type, collaborators, countries, number of sites, eligibility basics, primary outcomes and whether… License: other.
+- [Trialtide: every ClinicalTrials.gov study in one table](https://huggingface.co/datasets/shaw276/trialtide-clinicaltrials-registry): All 604,146 studies registered on ClinicalTrials.gov, flattened to one row per study: status, phase, design, enrollment, key dates, conditions (with MeSH terms), interventions, lead sponsor and sponsor type, collaborators, countries, number of sites, eligibility basics, primary outcomes and whether… License: other (us-public-domain-clinicaltrials-gov).
 - [Fipsbook: US Census Geographies with FIPS Codes, Centroids & Land Area (2025)](https://huggingface.co/datasets/shaw276/fipsbook-us-census-geographies): Free lookup tables for joining, mapping and geocoding US data: 210,319 rows across 11 layers (states, 3,222 counties, 32,350 places, 36,427 county subdivisions, 33,791 ZIP code areas (ZCTAs), 85,396 Census tracts, 935 metro/micro areas, 440 congressional districts of the 119th Congress, 6,843 state… License: cc0-1.0.
-- [Toolroll: Official MCP Registry Snapshot (every MCP server, weekly)](https://huggingface.co/datasets/shaw276/toolroll-mcp-registry-snapshot): A snapshot of the official Model Context Protocol (MCP) Registry (registry.modelcontextprotocol.io) as a single Parquet table: 35,403 MCP servers (35,028 active, 375 deprecated) from 20,000+ publishers, snapshot 2026-09-24 07:03 UTC. License: other.
+- [Toolroll: Official MCP Registry Snapshot (every MCP server, weekly)](https://huggingface.co/datasets/shaw276/toolroll-mcp-registry-snapshot): A snapshot of the official Model Context Protocol (MCP) Registry (registry.modelcontextprotocol.io) as a single Parquet table: 35,403 MCP servers (35,028 active, 375 deprecated) from 20,000+ publishers, snapshot 2026-09-24 07:03 UTC. License: other (mcp-registry-public-metadata).
 - [Tallyfold: SEC XBRL Financial Statements & Key Metrics (every US filer, latest 4 quarters)](https://huggingface.co/datasets/shaw276/tallyfold-sec-xbrl-financials): The numbers from every financial statement filed with the SEC in the latest four quarters: 10-K, 10-Q, 20-F and 40-F reports from 7,000+ US-listed and SEC-registered companies, as tidy parquet you can query in one line. License: cc0-1.0.
 - [Pingroll: MCP Server Uptime (daily handshake probes of every remote server in the official MCP Registry)](https://huggingface.co/datasets/shaw276/pingroll-mcp-server-uptime): Registries and directories tell you an MCP server exists. This dataset tells you whether it works: every public remote server in the official MCP Registry gets a real MCP handshake once a day (initialize → notifications/initialized → tools/list; no tool is ever called), and each result is recorded… License: cc0-1.0.
 - [Riskroll: SEC 10-K & 10-Q Risk Factors, MD&A and Market Risk Text (clean, by section)](https://huggingface.co/datasets/shaw276/riskroll-sec-10k-10q-sections): The parts of annual and quarterly reports that analysts, researchers and LLM pipelines actually read, cut out of each filing and cleaned: Risk Factors (Item 1A), Management's Discussion and Analysis (MD&A), Quantitative and Qualitative Disclosures About Market Risk, Cybersecurity (Item 1C) and… License: cc0-1.0.
 - [Policyloom: US AI Policy Tracker (Federal Register, daily)](https://huggingface.co/datasets/shaw276/policyloom-us-ai-policy-tracker): Every document in the US Federal Register since 2017-01-01 whose text mentions artificial intelligence: executive orders and other presidential documents, final rules, proposed rules, notices, requests for information and their comment periods, with dates, agencies, dockets and links. License: cc0-1.0.
-- [Swellmeter: Google Trending Searches (daily snapshot, 30 countries)](https://huggingface.co/datasets/shaw276/swellmeter-google-trending-searches): A daily snapshot of Google Trends' "Trending now" list for 30 countries: each trending search, its approximate traffic bucket, when it started trending, and the news articles Google attached to it. License: other.
-- [HireHeat Jobs: career-site job postings, first posted in the last 7 days (weekly snapshots)](https://huggingface.co/datasets/shaw276/hireheat-career-site-jobs-weekly): Each snapshot holds every job first posted in the last 7 days on the company career sites in the HireHeat Jobs directory, read straight from each employer's applicant tracking system (Greenhouse, Ashby, Workable, SmartRecruiters). License: other.
+- [Swellmeter: Google Trending Searches (daily snapshot, 30 countries)](https://huggingface.co/datasets/shaw276/swellmeter-google-trending-searches): A daily snapshot of Google Trends' "Trending now" list for 30 countries: each trending search, its approximate traffic bucket, when it started trending, and the news articles Google attached to it. License: other (google-trends-public-feed).
+- [HireHeat Jobs: career-site job postings, first posted in the last 7 days (weekly snapshots)](https://huggingface.co/datasets/shaw276/hireheat-career-site-jobs-weekly): Each snapshot holds every job first posted in the last 7 days on the company career sites in the HireHeat Jobs directory, read straight from each employer's applicant tracking system (Greenhouse, Ashby, Workable, SmartRecruiters). License: other (public-job-postings).
 - [Gripeline: SmartThings App Review Mood Around the Sept 2026 Fridge Firmware Incident](https://huggingface.co/datasets/shaw276/smartthings-review-mood-sept-2026): On 22 Sep 2026 a refrigerator firmware update pushed through Samsung's SmartThings platform left some smart fridges in Korea without power (Ars Technica, 23 Sep 2026; Samsung told Ars the issue was limited to Korea and came from an error during internal testing). License: cc-by-4.0.
 - [Keyfern: What People Search About Superintelligence (Sept 2026)](https://huggingface.co/datasets/shaw276/superintelligence-search-questions): "Superintelligence" moved from books and sci-fi into the news in the week of 22 Sep 2026: an AI remark in a UN General Assembly speech (22 Sep) and a proposed US bill to ban artificial superintelligence (reported by AP on 23 Sep; see the Policyloom AI policy tracker for the verified events). License: cc-by-4.0.
 - [Policyloom: Superintelligence Ban Bill, News Coverage & Search Interest (Sept 2026)](https://huggingface.co/datasets/shaw276/superintelligence-ban-bill-coverage): In September 2026 "superintelligence" became a US policy topic: a US senator and a House member announced a bill to ban the development of artificial superintelligence and pause advanced AI development until a federal regulator sets safety rules (Senate press release, 3 Sep; widely reported on 23… License: cc-by-4.0.
 
+## Web apps
+
+- [Sumbloom — Free Daily Number Puzzle (Math Game Like Digits)](https://shaw276-cybermax.static.hf.space/sumbloom/index.html): Free daily number puzzle: combine six numbers with + − × ÷ to hit five targets. Free, for people (no API).
+- [Honeyfold — Free Idle Bee Game You Can Finish](https://shaw276-cybermax.static.hf.space/honeyfold/index.html): Honeyfold is a free idle clicker game: grow a bee colony from one bee to the Golden Comb in about 30 minutes. Free, for people (no API).
+- [Free UTM Builder — Bulk Campaign URL Builder + QR Code](https://shaw276-cybermax.static.hf.space/utm-builder/index.html): Free UTM builder: tag campaign URLs for Google Analytics 4, build hundreds at once from a list or CSV, make a QR code, and decode tagged links. Free, for people (no API).
+- [Salary to Hourly Calculator + Freelance Rate Calculator](https://shaw276-cybermax.static.hf.space/salary-to-hourly-calculator/index.html): Free salary to hourly calculator: convert annual pay to hourly, daily, weekly and monthly, hourly to salary with overtime, and work out your freelance rate. Free, for people (no API).
+- [Profit Margin Calculator + Markup Calculator (Free)](https://shaw276-cybermax.static.hf.space/profit-margin-calculator/index.html): Free profit margin calculator: enter any two of cost, price, profit, margin or markup and get the rest. Free, for people (no API).
+- [Hushscribe — Private Audio Transcription](https://shaw276-hushscribe.static.hf.space/index.html): Free private audio and video transcription that runs in your browser with Whisper. Free, for people (no API).
+
 ## Optional
 
+- [Fieldwright (coming soon, not callable yet): AI Web Data Extractor (Bring Your Own LLM Key)](https://shaw276-cybermax.static.hf.space/market/tools/ai-extract.md): AI web scraper that turns any page into typed JSON fields you name, using YOUR OpenAI, Claude, Gemini or OpenRouter key (you pay the model at cost).
 - [Inkpluck (coming soon, not callable yet): Article Extractor: URL to Text, Markdown & Metadata](https://shaw276-cybermax.static.hf.space/market/tools/article-extract.md): Article extractor API for news, blogs and docs: URLs or HTML in, clean text + Markdown, title, authors, publish date, language, top image, videos and paywall flag out.
 - [Pinloft (coming soon, not callable yet): Batch Geocoder: Address to Lat/Long + Census FIPS](https://shaw276-cybermax.static.hf.space/market/tools/batch-geocoder.md): Batch geocoding API: addresses to latitude/longitude, cleaned address, county, Census tract and block FIPS for US addresses (US Census), worldwide via OpenStreetMap.
 - [Picpail (coming soon, not callable yet): Bulk Image Downloader: All Images from URLs to ZIP](https://shaw276-cybermax.static.hf.space/market/tools/bulk-image-downloader.md): Download all images from web pages or image URLs into one ZIP, full-size originals instead of thumbnails: img, srcset, lazy-load, CSS, og:image.
 - [Firmfold (coming soon, not callable yet): Company Enrichment API: Logo, Tech, Contacts, Jobs](https://shaw276-cybermax.static.hf.space/market/tools/company-enrichment.md): Company enrichment API: give domains, work emails or company names; get one flat profile each: official domain, logo + brand colour, tech stack, email security, domain age, contacts, socials, hiring and Wikidata firmographics.
 - [Namewhere (coming soon, not callable yet): Company Name to Domain & Website Finder](https://shaw276-cybermax.static.hf.space/market/tools/company-to-domain.md): Company name to domain finder: turn company names into official website domains with a confidence score and evidence (Wikidata, homepage brand check), alternatives, social links, parked domains rejected.
 - [Commonkite (coming soon, not callable yet): Free Image Search API: Creative Commons & CC0](https://shaw276-cybermax.static.hf.space/market/tools/creative-commons-images.md): Search free-to-use images by keyword across Openverse (900M+ CC images), NASA, The Met, Art Institute of Chicago, Cleveland Museum of Art and Wikimedia Commons.
+- [Bonafido (coming soon, not callable yet): Business Domain Verification API](https://shaw276-cybermax.static.hf.space/market/tools/domain-verify.md): Checks whether a domain, URL or email domain is a real, operating, legitimate business website and (optionally) whether it belongs to the company you expect; returns a verdict, a 0-100 trust score, machine-readable risk flags with reasons, and quotable evidence.
+- [Mailvett (coming soon, not callable yet): Email Validation API for AI Agents](https://shaw276-cybermax.static.hf.space/market/tools/email-validate.md): Validate email addresses before sending or saving them: syntax, domain and MX / mail provider, disposable, role, free-provider and typo flags, domain age, SPF/DMARC and parked-domain signals, returned as a verdict (deliverable_likely, risky, undeliverable, invalid) with reason codes and a 0-100…
 - [Printwren (coming soon, not callable yet): HTML to PDF & URL to PDF API](https://shaw276-cybermax.static.hf.space/market/tools/html-to-pdf.md): HTML to PDF and URL to PDF API with headless Chrome: A4/Letter/Legal, landscape, margins, header/footer with page numbers, print CSS.
 - [Logolark (coming soon, not callable yet): Company Logo API & Favicon Finder by Domain](https://shaw276-cybermax.static.hf.space/market/tools/logo-finder.md): Clearbit Logo API replacement: give domains, URLs or emails, get each company's own best logo and square icon, verified by download: real size, SVG/PNG/ICO, transparency, brand colour, hosted copy, plus rebrand alerts in monitor mode.
 - [Paperkiln (coming soon, not callable yet): Word, Excel & PowerPoint to PDF Converter API](https://shaw276-cybermax.static.hf.space/market/tools/office-to-pdf.md): Convert DOCX, XLSX, PPTX, ODT, RTF, CSV and legacy DOC/XLS/PPT files to PDF (or PDF/A) in bulk with headless LibreOffice and Office-compatible fonts.

market/mcp/ai-extract.tools.json

@@ -0,0 +1,351 @@
+{
+  "$comment": "MCP tools/list result shape (MCP spec 2025-06-18) for Fieldwright. Generated by scripts/market/build.mjs.",
+  "server": {
+    "name": "cybermax-ai-extract",
+    "url": "https://mcp.apify.com/?tools=cybermax/ai-extract",
+    "transport": "streamable-http",
+    "auth": "Authorization: Bearer <APIFY_TOKEN>",
+    "local": null
+  },
+  "tools": [
+    {
+      "name": "cybermax--ai-extract",
+      "title": "Fieldwright: AI Web Data Extractor (Bring Your Own LLM Key)",
+      "description": "AI web scraper that turns any page into typed JSON fields you name, using YOUR OpenAI, Claude, Gemini or OpenRouter key (you pay the model at cost). List mode, change monitoring, JSON-LD. $0.001/page; failures free. Pricing: Page extracted with your LLM $0.001; Page parsed without LLM $0.0003 (billed to the caller's Apify account).",
+      "inputSchema": {
+        "type": "object",
+        "properties": {
+          "urls": {
+            "type": "array",
+            "title": "Page URLs",
+            "description": "Pages to extract from (product pages, listings, company sites, job posts, events...). One URL per line; the scheme is optional. Duplicates are skipped.",
+            "examples": [
+              [
+                "https://books.toscrape.com/catalogue/a-light-in-the-attic_1000/index.html",
+                "https://books.toscrape.com/catalogue/tipping-the-velvet_999/index.html"
+              ]
+            ],
+            "items": {
+              "type": "string"
+            }
+          },
+          "fields": {
+            "type": "array",
+            "title": "Fields to extract",
+            "description": "The columns you want. Each: {\"name\": \"price\", \"type\": \"number\", \"description\": \"Price in GBP\"}. Types: string, number, integer, boolean, url, email, date, string[], number[]. A plain list of names also works: [\"title\", \"price\"]. Descriptions steer the model; values not on the page come back as null (never guessed).",
+            "examples": [
+              [
+                {
+                  "name": "title",
+                  "type": "string",
+                  "description": "Book title"
+                },
+                {
+                  "name": "price",
+                  "type": "number",
+                  "description": "Price including tax, number only"
+                },
+                {
+                  "name": "in_stock",
+                  "type": "boolean"
+                },
+                {
+                  "name": "stock_count",
+                  "type": "integer",
+                  "description": "Number of copies available"
+                },
+                {
+                  "name": "upc",
+                  "type": "string"
+                },
+                {
+                  "name": "category",
+                  "type": "string",
+                  "description": "Book category from the breadcrumb"
+                }
+              ]
+            ],
+            "items": {}
+          },
+          "instructions": {
+            "type": "string",
+            "title": "Extra instructions (optional)",
+            "description": "Plain-English guidance for the model, e.g. \"Prices in USD without the currency sign\" or \"Only include speakers, not moderators\"."
+          },
+          "extractList": {
+            "type": "boolean",
+            "title": "List mode: many items per page",
+            "description": "On: every matching item on each page becomes its own row (products on a category page, jobs on a careers page, people on a team page). Still charged once per page.",
+            "default": false
+          },
+          "jsonSchema": {
+            "type": "object",
+            "title": "Advanced: your own JSON Schema (optional)",
+            "description": "Instead of Fields, give an object JSON Schema; its top-level properties become the fields (types and descriptions kept)."
+          },
+          "llmProvider": {
+            "type": "string",
+            "title": "LLM provider (your own key)",
+            "description": "Fieldwright calls the model with YOUR key, so you pay the provider directly at their price and choose the model. With \"none\", you get the no-LLM record (title, description, OpenGraph, JSON-LD, emails, phones, Markdown) at a lower price.",
+            "default": "none",
+            "enum": [
+              "none",
+              "openai",
+              "anthropic",
+              "gemini",
+              "openrouter",
+              "openai-compatible"
+            ],
+            "x-enumTitles": [
+              "None: metadata, JSON-LD and page Markdown only (no key needed)",
+              "OpenAI",
+              "Anthropic (Claude)",
+              "Google Gemini",
+              "OpenRouter (hundreds of models)",
+              "Any OpenAI-compatible API (Groq, Together, DeepSeek, Mistral, local...)"
+            ]
+          },
+          "llmApiKey": {
+            "type": "string",
+            "title": "Your API key",
+            "description": "Stored encrypted by Apify as a secret input. It is only sent to the provider you picked, never logged or saved in the dataset.",
+            "writeOnly": true
+          },
+          "llmModel": {
+            "type": "string",
+            "title": "Model (optional)",
+            "description": "Leave empty for a fast, low-cost default (OpenAI gpt-4.1-mini, Claude claude-haiku-4-5, Gemini gemini-2.5-flash, OpenRouter openai/gpt-4.1-mini) or type any model your key can use."
+          },
+          "llmBaseUrl": {
+            "type": "string",
+            "title": "Base URL (OpenAI-compatible / proxies only)",
+            "description": "For \"Any OpenAI-compatible API\", e.g. https://api.groq.com/openai/v1. Leave empty otherwise."
+          },
+          "maxOutputTokens": {
+            "type": "integer",
+            "title": "Max output tokens per page",
+            "description": "Default 2048 (8192 in list mode). Raise for pages with many items.",
+            "minimum": 256,
+            "maximum": 32000
+          },
+          "maxPageChars": {
+            "type": "integer",
+            "title": "Max page characters sent to the model",
+            "description": "Caps tokens (and your LLM cost) per page. The page is converted to compact Markdown with links, tables and JSON-LD first.",
+            "default": 20000,
+            "minimum": 2000,
+            "maximum": 200000
+          },
+          "onlyChanged": {
+            "type": "boolean",
+            "title": "Monitor: output only pages whose data changed",
+            "description": "Remembers each page's extracted data between runs with the same monitor name. Unchanged pages are not output and not charged; rows get changed = new or updated.",
+            "default": false
+          },
+          "monitorName": {
+            "type": "string",
+            "title": "Monitor name",
+            "description": "Separate memory per watch list, e.g. \"competitor-prices\"."
+          },
+          "includeMarkdown": {
+            "type": "boolean",
+            "title": "Include page Markdown in LLM rows",
+            "description": "Adds the Markdown the model read (useful for audits and RAG).",
+            "default": false
+          },
+          "includeMetadata": {
+            "type": "boolean",
+            "title": "Include metadata in LLM rows",
+            "description": "Adds description, language, canonical URL, OpenGraph, JSON-LD, emails and phones next to your fields.",
+            "default": false
+          },
+          "htmlDocuments": {
+            "type": "array",
+            "title": "Raw HTML instead of URLs (optional)",
+            "description": "[{\"url\": \"https://example.com/page\", \"html\": \"<html>...\"}] for pages you already fetched.",
+            "items": {}
+          },
+          "maxPages": {
+            "type": "integer",
+            "title": "Max pages",
+            "description": "Stop after this many pages (a safety cap on your LLM spend).",
+            "default": 1000,
+            "minimum": 1,
+            "maximum": 10000
+          },
+          "maxConcurrency": {
+            "type": "integer",
+            "title": "Parallel pages",
+            "description": "Lower it if your LLM key has a small rate limit.",
+            "default": 5,
+            "minimum": 1,
+            "maximum": 20
+          },
+          "timeoutSecs": {
+            "type": "integer",
+            "title": "Page load timeout (s)",
+            "description": "How long to wait for each page to load.",
+            "default": 25,
+            "minimum": 5,
+            "maximum": 120
+          },
+          "retryBlockedWithProxy": {
+            "type": "boolean",
+            "title": "Retry blocked pages through Apify Proxy",
+            "description": "If a site answers 401/403/429/503 to a direct request, try once more through Apify Proxy.",
+            "default": true
+          },
+          "proxyConfiguration": {
+            "type": "object",
+            "title": "Proxy (optional)",
+            "description": "Send every page request through this proxy. LLM calls never go through it.",
+            "default": {
+              "useApifyProxy": false
+            }
+          },
+          "waitSecs": {
+            "type": "integer",
+            "description": "Optional (hosted Apify MCP only): seconds to wait for the run before returning a run handle."
+          }
+        },
+        "title": "Fieldwright AI web data extractor",
+        "additionalProperties": false
+      },
+      "outputSchema": {
+        "type": "object",
+        "properties": {
+          "items": {
+            "type": "array",
+            "items": {
+              "type": "object",
+              "properties": {
+                "url": {
+                  "type": "string",
+                  "description": "URL as given"
+                },
+                "finalUrl": {
+                  "type": [
+                    "string",
+                    "null"
+                  ]
+                },
+                "ok": {
+                  "type": "boolean",
+                  "description": "true = processed (charged once per page); false = free failure, see error"
+                },
+                "mode": {
+                  "type": "string",
+                  "description": "single, list or no-llm"
+                },
+                "title": {
+                  "type": [
+                    "string",
+                    "null"
+                  ]
+                },
+                "data": {
+                  "type": [
+                    "object",
+                    "null"
+                  ],
+                  "description": "Your fields, typed; null where the page doesn't say"
+                },
+                "missingFields": {
+                  "type": "array",
+                  "items": {
+                    "type": "string"
+                  }
+                },
+                "itemIndex": {
+                  "type": [
+                    "integer",
+                    "null"
+                  ]
+                },
+                "itemsFound": {
+                  "type": "integer"
+                },
+                "changed": {
+                  "type": "string",
+                  "description": "Monitor mode: new or updated"
+                },
+                "llm": {
+                  "type": "object",
+                  "description": "provider, model, inputTokens, outputTokens (billed to your key)"
+                },
+                "description": {
+                  "type": [
+                    "string",
+                    "null"
+                  ]
+                },
+                "language": {
+                  "type": [
+                    "string",
+                    "null"
+                  ]
+                },
+                "canonicalUrl": {
+                  "type": [
+                    "string",
+                    "null"
+                  ]
+                },
+                "openGraph": {
+                  "type": "object"
+                },
+                "jsonLd": {
+                  "type": "array"
+                },
+                "emails": {
+                  "type": "array"
+                },
+                "phones": {
+                  "type": "array"
+                },
+                "markdown": {
+                  "type": [
+                    "string",
+                    "null"
+                  ]
+                },
+                "pageChars": {
+                  "type": "integer"
+                },
+                "truncated": {
+                  "type": "boolean"
+                },
+                "httpStatus": {
+                  "type": [
+                    "integer",
+                    "null"
+                  ]
+                },
+                "fetchedVia": {
+                  "type": "string"
+                },
+                "ms": {
+                  "type": "integer"
+                },
+                "error": {
+                  "type": "string"
+                },
+                "extractedAt": {
+                  "type": "string"
+                }
+              }
+            }
+          }
+        }
+      },
+      "annotations": {
+        "title": "Fieldwright",
+        "readOnlyHint": true,
+        "destructiveHint": false,
+        "idempotentHint": false,
+        "openWorldHint": true
+      }
+    }
+  ],
+  "localTools": []
+}

market/mcp/creative-commons-images.tools.json

@@ -5,13 +5,13 @@
     "url": "https://mcp.apify.com/?tools=cybermax/creative-commons-images",
     "transport": "streamable-http",
     "auth": "Authorization: Bearer <APIFY_TOKEN>",
-    "local": null
+    "local": "mcp/creative-commons-images-mcp"
   },
   "tools": [
     {
       "name": "cybermax--creative-commons-images",
       "title": "Commonkite: Free Image Search API: Creative Commons & CC0",
-      "description": "Search free-to-use images by keyword across Openverse (900M+ CC images), NASA, The Met, Art Institute of Chicago, Cleveland Museum of Art and Wikimedia Commons. Every image comes with its license, commercial-use and modification flags, and a ready-to-paste attribution line. $0.001 per image. Pricing: Image found $0.001 (billed to the caller's Apify account).",
+      "description": "Commonkite. Free image search by keyword across Openverse (900M+ Creative Commons images from Flickr, Wikimedia and museums), NASA, The Met, Art Institute of Chicago, Cleveland Museum of Art and optionally Wikimedia Commons live search. Returns, per image: imageUrl, thumbnailUrl, width, height, title, creator, landingUrl, source, licenseName and licenseUrl, the flags commercialUse, modificationsAllowed and attributionRequired, and attributionText (a ready-to-paste credit line). License filter: \"commercial\" (default), \"commercial-modify\", \"no-attribution\" (CC0/public domain only) or \"any\". mode=\"random\" returns one random image per term (placeholder/hero images). Use it to find a hero image for a blog post you are writing, illustrate slides with properly credited photos, or pick CC0 images that need no credit. Not for photos of specific products, logos, celebrities or today's news (rarely openly licensed); licenses are reported as the source states them. Cost: $0.001 per image returned; searches with no results are free. Pricing: Image found $0.001 (billed to the caller's Apify account).",
       "inputSchema": {
         "type": "object",
         "properties": {
@@ -355,5 +355,11 @@
       }
     }
   ],
-  "localTools": []
+  "localTools": [
+    {
+      "name": "search_free_images",
+      "title": "Commonkite: search free-to-use Creative Commons & CC0 images",
+      "description": "Commonkite. Free image search by keyword across Openverse (900M+ Creative Commons images from Flickr, Wikimedia and museums), NASA, The Met, Art Institute of Chicago, Cleveland Museum of Art and optionally Wikimedia Commons live search. Returns, per image: imageUrl, thumbnailUrl, width, height, title, creator, landingUrl, source, licenseName and licenseUrl, the flags commercialUse, modificationsAllowed and attributionRequired, and attributionText (a ready-to-paste credit line). License filter: \"commercial\" (default), \"commercial-modify\", \"no-attribution\" (CC0/public domain only) or \"any\". mode=\"random\" returns one random image per term (placeholder/hero images). Use it to find a hero image for a blog post you are writing, illustrate slides with properly credited photos, or pick CC0 images that need no credit. Not for photos of specific products, logos, celebrities or today's news (rarely openly licensed); licenses are reported as the source states them. Cost: $0.001 per image returned; searches with no results are free."
+    }
+  ]
 }

market/mcp/cybermax-datasets.tools.json

@@ -0,0 +1,99 @@
+{
+  "$comment": "MCP tools/list result shape (MCP spec 2025-06-18) for the free CyberMax datasets server. Generated by scripts/market/build.mjs.",
+  "server": {
+    "name": "cybermax-datasets-mcp",
+    "url": null,
+    "transport": "stdio",
+    "auth": "none",
+    "local": "mcp/cybermax-datasets-mcp"
+  },
+  "tools": [
+    {
+      "name": "list_datasets",
+      "title": "CyberMax datasets: list",
+      "description": "Lists every CyberMax open dataset with a one-line summary, licence and the related paid tool.",
+      "inputSchema": {
+        "type": "object",
+        "properties": {}
+      },
+      "annotations": {
+        "title": "CyberMax datasets: list",
+        "readOnlyHint": true,
+        "destructiveHint": false,
+        "idempotentHint": true,
+        "openWorldHint": true
+      }
+    },
+    {
+      "name": "describe_dataset",
+      "title": "CyberMax datasets: columns and sample rows",
+      "description": "Configs/splits, column names and types, and 3 sample rows of one dataset.",
+      "inputSchema": {
+        "type": "object",
+        "properties": {
+          "dataset": {
+            "type": "string",
+            "description": "Dataset name from list_datasets."
+          }
+        },
+        "required": [
+          "dataset"
+        ]
+      },
+      "annotations": {
+        "title": "CyberMax datasets: columns and sample rows",
+        "readOnlyHint": true,
+        "destructiveHint": false,
+        "idempotentHint": true,
+        "openWorldHint": true
+      }
+    },
+    {
+      "name": "query_dataset",
+      "title": "CyberMax datasets: search, filter or page rows",
+      "description": "Full-text search, SQL-style where/orderby filter (DuckDB syntax) or plain paging over a dataset; up to 100 rows per call plus the total match count.",
+      "inputSchema": {
+        "type": "object",
+        "required": [
+          "dataset"
+        ],
+        "properties": {
+          "dataset": {
+            "type": "string"
+          },
+          "query": {
+            "type": "string"
+          },
+          "where": {
+            "type": "string"
+          },
+          "orderby": {
+            "type": "string"
+          },
+          "config": {
+            "type": "string"
+          },
+          "split": {
+            "type": "string"
+          },
+          "offset": {
+            "type": "integer",
+            "minimum": 0
+          },
+          "length": {
+            "type": "integer",
+            "minimum": 1,
+            "maximum": 100
+          }
+        }
+      },
+      "annotations": {
+        "title": "CyberMax datasets: search, filter or page rows",
+        "readOnlyHint": true,
+        "destructiveHint": false,
+        "idempotentHint": true,
+        "openWorldHint": true
+      }
+    }
+  ]
+}

market/mcp/domain-verify.tools.json

@@ -0,0 +1,532 @@
+{
+  "$comment": "MCP tools/list result shape (MCP spec 2025-06-18) for Bonafido. Generated by scripts/market/build.mjs.",
+  "server": {
+    "name": "io.github.shawmena-dotcom/domain-verify",
+    "url": "https://mcp.apify.com/?tools=cybermax/domain-verify",
+    "transport": "streamable-http",
+    "auth": "Authorization: Bearer <APIFY_TOKEN>",
+    "local": "mcp/domain-verify-mcp"
+  },
+  "tools": [
+    {
+      "name": "cybermax--domain-verify",
+      "title": "Bonafido: Business Domain Verification API",
+      "description": "Bonafido. For each domain, URL or email (optionally with the company name you expect) returns one JSON record: verdict (verified | likely_legit | unverified | suspicious | not_found), a 0-100 trustScore, riskFlags (parked, new_domain_lt_30d/lt_90d, lookalike_of:<domain>, name_mismatch, not_official_domain, no_website, placeholder_page, tls_invalid, free_email_domain, redirects_offsite, no_mx, no_dmarc, no_contact_info) each with a reason, and evidence strings with their source (RDAP domain age, TLS, MX/SPF/DMARC, homepage, contact page/email/phone/address, social profiles, Wikidata official website, company-name match). Use it to decide whether a vendor, payee, lead or sign-up website is a real operating business and whether it belongs to the named company before paying, onboarding or emailing. Signals only, never a fraud finding. For raw tech-stack, DNS, security-header or WHOIS data use DomainDNA analyze_domains instead. Costs $0.006 per existing domain plus $0.004 when a company name is checked, billed to your Apify account; invalid and non-existent domains are free. About 1-3 s per domain, no browser. Pricing: Domain verified $0.006; Company name checked $0.004 (billed to the caller's Apify account).",
+      "inputSchema": {
+        "type": "object",
+        "properties": {
+          "items": {
+            "type": "array",
+            "title": "Domains, URLs or emails to verify",
+            "description": "One entry per website to check. Plain strings (\"acme.com\", \"https://www.acme.com/about\", \"billing@acme.com\") or objects with the company you expect: {\"domain\": \"stripe.com\", \"companyName\": \"Stripe\", \"country\": \"us\"}. The bar form \"acme.com | Acme Corp | us\" works too. With a company name you also get a name match against the domain, the homepage and Wikidata's official website, plus look-alike detection.",
+            "examples": [
+              [
+                {
+                  "domain": "stripe.com",
+                  "companyName": "Stripe"
+                }
+              ]
+            ],
+            "items": {}
+          },
+          "country": {
+            "type": "string",
+            "title": "Default country (ISO code)",
+            "description": "Optional two-letter code (\"us\", \"gb\", \"de\") applied to items without their own. Used to pick the right company on Wikidata when names are shared across countries."
+          },
+          "checkContacts": {
+            "type": "boolean",
+            "title": "Look for contact details",
+            "description": "Read the homepage and up to 'Contact pages to read' contact/about pages for a contact page, form, email, phone, street address and company social profiles.",
+            "default": true
+          },
+          "maxContactPages": {
+            "type": "integer",
+            "title": "Contact pages to read",
+            "description": "Extra pages per site (contact, about, imprint) read for contact signals. 0 = homepage only (faster).",
+            "default": 3,
+            "minimum": 0,
+            "maximum": 6
+          },
+          "useWikidata": {
+            "type": "boolean",
+            "title": "Cross-check Wikidata",
+            "description": "Look up the company (or the brand the site claims) on Wikidata to find its official website: confirms real companies and powers look-alike detection. Wikidata outages are noted, never counted against a domain.",
+            "default": true
+          },
+          "timeoutSecs": {
+            "type": "integer",
+            "title": "Per-request timeout (seconds)",
+            "description": "Timeout for each HTTP, DNS, TLS or RDAP request.",
+            "default": 10,
+            "minimum": 3,
+            "maximum": 60
+          },
+          "maxConcurrency": {
+            "type": "integer",
+            "title": "Items in parallel",
+            "description": "How many items to verify at the same time.",
+            "default": 5,
+            "minimum": 1,
+            "maximum": 20
+          },
+          "monitor": {
+            "type": "boolean",
+            "title": "Monitor mode: only return changes",
+            "description": "Watch this list over time (vendors, payees, partners). Each run compares every item with the previous run of the same watchlist and returns (and charges) only items that are new or whose verdict, trust band (10-point steps) or risk flags changed, e.g. a supplier's domain got parked, its certificate broke or it stopped receiving mail. Unchanged items are skipped and not charged. Made for daily or weekly Schedules.",
+            "default": false
+          },
+          "monitorKey": {
+            "type": [
+              "string",
+              "null"
+            ],
+            "title": "Watchlist name (monitor mode)",
+            "description": "Optional name for this watchlist's saved state, e.g. \"vendors-eu\", so several watchlists stay separate and you can edit the list without losing history. Empty = derived from the list."
+          },
+          "waitSecs": {
+            "type": "integer",
+            "description": "Optional (hosted Apify MCP only): seconds to wait for the run before returning a run handle."
+          }
+        },
+        "title": "Bonafido: business domain verification",
+        "required": [
+          "items"
+        ],
+        "additionalProperties": false
+      },
+      "outputSchema": {
+        "type": "object",
+        "properties": {
+          "items": {
+            "type": "array",
+            "items": {
+              "type": "object",
+              "properties": {
+                "input": {
+                  "description": "The item as you sent it (string or object).",
+                  "type": [
+                    "string",
+                    "object",
+                    "null"
+                  ]
+                },
+                "domain": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "Hostname checked (www. removed, subdomains kept)."
+                },
+                "registrableDomain": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "Registrable domain used for DNS, mail and RDAP (e.g. acme.co.uk)."
+                },
+                "companyName": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "Company name you expected, if given."
+                },
+                "country": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "Two-letter country used for the Wikidata match."
+                },
+                "verdict": {
+                  "type": "string",
+                  "description": "verified | likely_legit | unverified | suspicious | not_found. Signals only, never a fraud finding.",
+                  "enum": [
+                    "verified",
+                    "likely_legit",
+                    "unverified",
+                    "suspicious",
+                    "not_found"
+                  ]
+                },
+                "trustScore": {
+                  "type": "integer",
+                  "description": "0-100 weighted trust score (website 25, domain age 20, TLS 10, email auth 12, contacts 13, socials 5, identity 15), capped by severe flags.",
+                  "minimum": 0,
+                  "maximum": 100
+                },
+                "summary": {
+                  "type": "string",
+                  "description": "One line an agent can quote."
+                },
+                "riskFlags": {
+                  "type": "array",
+                  "items": {
+                    "type": "string"
+                  },
+                  "description": "Machine-readable flags, most severe first: parked, registration_expired, registry_hold, new_domain_lt_30d, new_domain_lt_90d, lookalike_of:<domain>, name_mismatch, not_official_domain, no_website, placeholder_page, tls_invalid, free_email_domain, redirects_offsite, no_https, no_mx, no_dmarc, no_contact_info, name_unconfirmed."
+                },
+                "riskDetails": {
+                  "type": "array",
+                  "description": "Each flag with severity (high/medium/low) and a plain-English reason.",
+                  "items": {
+                    "type": "object",
+                    "properties": {
+                      "flag": {
+                        "type": "string"
+                      },
+                      "severity": {
+                        "type": "string",
+                        "enum": [
+                          "high",
+                          "medium",
+                          "low"
+                        ]
+                      },
+                      "reason": {
+                        "type": "string"
+                      }
+                    },
+                    "required": [
+                      "flag",
+                      "severity",
+                      "reason"
+                    ]
+                  }
+                },
+                "evidence": {
+                  "type": "array",
+                  "description": "Quotable facts, each with its source (dns, rdap, tls, homepage, website, schema.org, wikidata, name-match).",
+                  "items": {
+                    "type": "object",
+                    "properties": {
+                      "source": {
+                        "type": "string"
+                      },
+                      "text": {
+                        "type": "string"
+                      }
+                    },
+                    "required": [
+                      "source",
+                      "text"
+                    ]
+                  }
+                },
+                "resolves": {
+                  "type": [
+                    "boolean",
+                    "null"
+                  ],
+                  "description": "The domain exists (DNS, mail, TLS or website answered)."
+                },
+                "reachable": {
+                  "type": "boolean",
+                  "description": "A homepage answered over HTTPS or HTTP."
+                },
+                "httpStatus": {
+                  "type": [
+                    "integer",
+                    "null"
+                  ],
+                  "description": "HTTP status of the homepage (after redirects)."
+                },
+                "finalUrl": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "Homepage URL after redirects."
+                },
+                "redirectsToOtherDomain": {
+                  "type": "boolean",
+                  "description": "The homepage ends up on a different registrable domain."
+                },
+                "redirectTarget": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "That other domain, if any."
+                },
+                "pageType": {
+                  "type": "string",
+                  "description": "normal | parked | placeholder | bot_protected | error_page | none.",
+                  "enum": [
+                    "normal",
+                    "parked",
+                    "placeholder",
+                    "bot_protected",
+                    "error_page",
+                    "none"
+                  ]
+                },
+                "siteName": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "Brand name the site states (schema.org, og:site_name or title)."
+                },
+                "domainAgeDays": {
+                  "type": [
+                    "integer",
+                    "null"
+                  ],
+                  "description": "Days since registration (RDAP). Null when the TLD has no public RDAP."
+                },
+                "registeredAt": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "Registration date (RDAP)."
+                },
+                "registrar": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "Registrar (RDAP)."
+                },
+                "tlsValid": {
+                  "type": [
+                    "boolean",
+                    "null"
+                  ],
+                  "description": "Certificate trusted for this hostname and not expired. Null = no HTTPS."
+                },
+                "tlsDaysLeft": {
+                  "type": [
+                    "integer",
+                    "null"
+                  ],
+                  "description": "Days until the certificate expires."
+                },
+                "tlsIssuer": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "Certificate issuer."
+                },
+                "mx": {
+                  "type": "boolean",
+                  "description": "Has MX records (can receive email)."
+                },
+                "mxProvider": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "Mail provider from MX (Google Workspace, Microsoft 365...)."
+                },
+                "spf": {
+                  "type": "boolean",
+                  "description": "Has an SPF record."
+                },
+                "dmarcPolicy": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "DMARC policy (none/quarantine/reject) or null when missing."
+                },
+                "contact": {
+                  "type": "object",
+                  "description": "Public contact signals.",
+                  "properties": {
+                    "contactPage": {
+                      "type": "boolean"
+                    },
+                    "contactPageUrl": {
+                      "type": [
+                        "string",
+                        "null"
+                      ]
+                    },
+                    "contactForm": {
+                      "type": "boolean"
+                    },
+                    "emailFound": {
+                      "type": "boolean"
+                    },
+                    "emails": {
+                      "type": "array",
+                      "items": {
+                        "type": "string"
+                      }
+                    },
+                    "phoneFound": {
+                      "type": "boolean"
+                    },
+                    "phones": {
+                      "type": "array",
+                      "items": {
+                        "type": "string"
+                      }
+                    },
+                    "addressFound": {
+                      "type": "boolean"
+                    },
+                    "address": {
+                      "type": [
+                        "string",
+                        "null"
+                      ]
+                    }
+                  }
+                },
+                "socials": {
+                  "type": "object",
+                  "description": "Company profile links found on the site, by platform (linkedin, x, facebook, instagram, youtube, github...).",
+                  "additionalProperties": {
+                    "type": "string"
+                  }
+                },
+                "nameMatch": {
+                  "type": "object",
+                  "description": "Company-name check (when a name is given) and Wikidata official-website cross-check.",
+                  "properties": {
+                    "checked": {
+                      "type": "boolean"
+                    },
+                    "score": {
+                      "type": [
+                        "number",
+                        "null"
+                      ]
+                    },
+                    "level": {
+                      "type": "string",
+                      "enum": [
+                        "strong",
+                        "partial",
+                        "weak",
+                        "mismatch",
+                        "not_checked"
+                      ]
+                    },
+                    "officialDomains": {
+                      "type": "array",
+                      "items": {
+                        "type": "string"
+                      }
+                    },
+                    "matchesOfficial": {
+                      "type": [
+                        "boolean",
+                        "null"
+                      ]
+                    },
+                    "wikidata": {
+                      "type": [
+                        "object",
+                        "null"
+                      ],
+                      "properties": {
+                        "id": {
+                          "type": "string"
+                        },
+                        "label": {
+                          "type": "string"
+                        },
+                        "description": {
+                          "type": [
+                            "string",
+                            "null"
+                          ]
+                        },
+                        "url": {
+                          "type": "string"
+                        }
+                      }
+                    }
+                  }
+                },
+                "lookalike": {
+                  "type": "object",
+                  "description": "Possible imitation of another (official) domain.",
+                  "properties": {
+                    "suspected": {
+                      "type": "boolean"
+                    },
+                    "of": {
+                      "type": [
+                        "string",
+                        "null"
+                      ]
+                    },
+                    "reason": {
+                      "type": [
+                        "string",
+                        "null"
+                      ]
+                    }
+                  }
+                },
+                "charged": {
+                  "type": "array",
+                  "items": {
+                    "type": "string"
+                  },
+                  "description": "Events charged for this item (domain-verified, company-name-checked); empty = free."
+                },
+                "errors": {
+                  "type": "array",
+                  "items": {
+                    "type": "string"
+                  },
+                  "description": "Per-check problems (e.g. a TLD without public RDAP, Wikidata unavailable)."
+                },
+                "change": {
+                  "type": "string",
+                  "description": "Monitor mode: new or changed."
+                },
+                "changes": {
+                  "type": "object",
+                  "description": "Monitor mode: verdict / trustBand from-to, riskFlags added/removed."
+                },
+                "checkedAt": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "ISO timestamp."
+                },
+                "ms": {
+                  "type": "integer",
+                  "description": "Time spent on this item (ms)."
+                }
+              },
+              "required": [
+                "domain",
+                "verdict",
+                "trustScore",
+                "summary",
+                "riskFlags",
+                "riskDetails",
+                "evidence"
+              ]
+            }
+          }
+        }
+      },
+      "annotations": {
+        "title": "Bonafido",
+        "readOnlyHint": true,
+        "destructiveHint": false,
+        "idempotentHint": false,
+        "openWorldHint": true
+      }
+    }
+  ],
+  "localTools": [
+    {
+      "name": "verify_business_domains",
+      "title": "Bonafido: verify business domains (is this a real, legit business website?)",
+      "description": "Bonafido. For each domain, URL or email (optionally with the company name you expect) returns one JSON record: verdict (verified | likely_legit | unverified | suspicious | not_found), a 0-100 trustScore, riskFlags (parked, new_domain_lt_30d/lt_90d, lookalike_of:<domain>, name_mismatch, not_official_domain, no_website, placeholder_page, tls_invalid, free_email_domain, redirects_offsite, no_mx, no_dmarc, no_contact_info) each with a reason, and evidence strings with their source (RDAP domain age, TLS, MX/SPF/DMARC, homepage, contact page/email/phone/address, social profiles, Wikidata official website, company-name match). Use it to decide whether a vendor, payee, lead or sign-up website is a real operating business and whether it belongs to the named company before paying, onboarding or emailing. Signals only, never a fraud finding. For raw tech-stack, DNS, security-header or WHOIS data use DomainDNA analyze_domains instead. Costs $0.006 per existing domain plus $0.004 when a company name is checked, billed to your Apify account; invalid and non-existent domains are free. About 1-3 s per domain, no browser."
+    }
+  ]
+}

market/mcp/email-validate.tools.json

@@ -0,0 +1,371 @@
+{
+  "$comment": "MCP tools/list result shape (MCP spec 2025-06-18) for Mailvett. Generated by scripts/market/build.mjs.",
+  "server": {
+    "name": "io.github.shawmena-dotcom/email-validate",
+    "url": "https://mcp.apify.com/?tools=cybermax/email-validate",
+    "transport": "streamable-http",
+    "auth": "Authorization: Bearer <APIFY_TOKEN>",
+    "local": "mcp/email-validate-mcp"
+  },
+  "tools": [
+    {
+      "name": "cybermax--email-validate",
+      "title": "Mailvett: Email Validation API for AI Agents",
+      "description": "Mailvett. Email validation / verification: give email addresses and get, per address, a VERDICT (deliverable_likely, risky, undeliverable, invalid), a 0-100 score and machine-readable reasons (e.g. mx_present, disposable_domain, role_account:sales, possible_typo_of:gmail.com, domain_not_found), plus the syntax check, MX hosts and mail provider (Gmail, Google Workspace, Microsoft 365, Zoho, Proton...), disposable, role, free-provider, typo and parked-domain flags, a corrected address when a typo is likely (gmial.com -> gmail.com), domain age (RDAP), SPF and DMARC. Use it to check an address BEFORE sending an email, saving a contact or accepting a sign-up, and to clean lead lists. It does not probe the mailbox over SMTP: mailbox and catchAll are always \"not_checked\". Cost: $0.0008 per address checked; invalid syntax and duplicates are free. About 1 s per new domain. Pricing: Email validated $0.0008 (billed to the caller's Apify account).",
+      "inputSchema": {
+        "type": "object",
+        "properties": {
+          "emails": {
+            "type": "array",
+            "title": "Email addresses",
+            "description": "One address per line (commas and semicolons also split). \"Jane Doe <jane@acme.com>\" and \"mailto:\" links are cleaned for you; duplicates are checked once. In JSON you can also pass objects: {\"email\": \"jane@acme.com\"}.",
+            "examples": [
+              [
+                "jane.doe@gmail.com",
+                "info@stripe.com",
+                "someone@mailinator.com",
+                "bob@gmial.com"
+              ]
+            ],
+            "items": {
+              "type": "string"
+            }
+          },
+          "suggestTypos": {
+            "type": "boolean",
+            "title": "Suggest typo fixes",
+            "description": "Flag likely misspellings of big mail providers (gmial.com → gmail.com, yahoo.con → yahoo.com) and return the corrected address in \"suggestion\".",
+            "default": true
+          },
+          "checkDomainAge": {
+            "type": "boolean",
+            "title": "Check domain age (RDAP)",
+            "description": "Look up the domain's registration date in the official RDAP registry. Very new domains (under 30 days) are a strong fraud / throwaway signal.",
+            "default": true
+          },
+          "checkWebsite": {
+            "type": "boolean",
+            "title": "Detect parked / for-sale domains",
+            "description": "Check the domain's name servers and homepage for parking or for-sale pages. Turn off for the fastest possible runs.",
+            "default": true
+          },
+          "maxConcurrency": {
+            "type": "integer",
+            "title": "Parallel checks",
+            "description": "Addresses checked at the same time. DNS answers are cached per domain within a run.",
+            "default": 10,
+            "minimum": 1,
+            "maximum": 50
+          },
+          "timeoutSecs": {
+            "type": "integer",
+            "title": "Timeout per lookup (seconds)",
+            "description": "Per DNS, RDAP or homepage request.",
+            "default": 8,
+            "minimum": 2,
+            "maximum": 60
+          },
+          "waitSecs": {
+            "type": "integer",
+            "description": "Optional (hosted Apify MCP only): seconds to wait for the run before returning a run handle."
+          }
+        },
+        "title": "Mailvett: email validation",
+        "required": [
+          "emails"
+        ],
+        "additionalProperties": false
+      },
+      "outputSchema": {
+        "type": "object",
+        "properties": {
+          "items": {
+            "type": "array",
+            "items": {
+              "type": "object",
+              "additionalProperties": true,
+              "required": [
+                "input",
+                "email",
+                "verdict",
+                "score",
+                "reasons",
+                "syntax",
+                "mx",
+                "flags",
+                "catchAll",
+                "mailbox",
+                "checkedAt"
+              ],
+              "properties": {
+                "input": {
+                  "type": "string",
+                  "description": "The address exactly as given."
+                },
+                "email": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "Normalised address (trimmed, display name and mailto removed, lower-cased, IDN domain in punycode)."
+                },
+                "verdict": {
+                  "type": "string",
+                  "enum": [
+                    "deliverable_likely",
+                    "risky",
+                    "undeliverable",
+                    "invalid"
+                  ],
+                  "description": "deliverable_likely = domain accepts mail and no risk signal; risky = disposable, parked, typo, very new domain or DNS failure; undeliverable = domain or mail server missing; invalid = bad syntax."
+                },
+                "score": {
+                  "type": "integer",
+                  "minimum": 0,
+                  "maximum": 100,
+                  "description": "0-100 confidence the address can receive mail. Capped at 90 because the mailbox itself is not probed."
+                },
+                "reasons": {
+                  "type": "array",
+                  "items": {
+                    "type": "string"
+                  },
+                  "description": "Machine-readable reason codes, e.g. mx_present, disposable_domain, role_account:sales, possible_typo_of:gmail.com, domain_not_found."
+                },
+                "summary": {
+                  "type": "string",
+                  "description": "One-line human/LLM readable summary."
+                },
+                "localPart": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "Local part"
+                },
+                "domain": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "Domain"
+                },
+                "syntax": {
+                  "type": "object",
+                  "properties": {
+                    "valid": {
+                      "type": "boolean"
+                    },
+                    "error": {
+                      "type": [
+                        "string",
+                        "null"
+                      ],
+                      "description": "Syntax error code, e.g. missing_at_sign, consecutive_dots."
+                    }
+                  },
+                  "description": "Syntax"
+                },
+                "domainExists": {
+                  "type": [
+                    "boolean",
+                    "null"
+                  ],
+                  "description": "false = NXDOMAIN; null = not checked or DNS failed."
+                },
+                "mx": {
+                  "type": "object",
+                  "properties": {
+                    "present": {
+                      "type": "boolean",
+                      "description": "A mail server exists (MX, or the A record per RFC 5321)."
+                    },
+                    "hosts": {
+                      "type": "array",
+                      "items": {
+                        "type": "string"
+                      },
+                      "description": "MX hosts by priority (max 10)."
+                    },
+                    "provider": {
+                      "type": [
+                        "string",
+                        "null"
+                      ],
+                      "description": "Mail provider from MX hosts: Gmail, Google Workspace, Microsoft 365, Outlook.com, Yahoo Mail, Zoho Mail, Proton Mail, iCloud Mail, Proofpoint, Mimecast, Other / self-hosted..."
+                    },
+                    "implicit": {
+                      "type": "boolean",
+                      "description": "No MX record; the domain's A/AAAA record is the implicit mail server."
+                    },
+                    "nullMx": {
+                      "type": "boolean",
+                      "description": "The domain publishes a null MX (RFC 7505): it accepts no mail."
+                    }
+                  },
+                  "description": "Mail server"
+                },
+                "flags": {
+                  "type": "object",
+                  "properties": {
+                    "disposable": {
+                      "type": [
+                        "boolean",
+                        "null"
+                      ],
+                      "description": "Domain is on the disposable / temporary email list."
+                    },
+                    "role": {
+                      "type": [
+                        "boolean",
+                        "null"
+                      ],
+                      "description": "Shared role mailbox (info@, sales@, support@...)."
+                    },
+                    "roleType": {
+                      "type": [
+                        "string",
+                        "null"
+                      ],
+                      "enum": [
+                        "general",
+                        "sales",
+                        "support",
+                        "press",
+                        "careers",
+                        "legal",
+                        "billing",
+                        "technical",
+                        "investors",
+                        "noreply",
+                        null
+                      ]
+                    },
+                    "freeProvider": {
+                      "type": [
+                        "boolean",
+                        "null"
+                      ],
+                      "description": "Consumer mailbox provider (Gmail, Outlook, Yahoo...)."
+                    },
+                    "typoSuggestion": {
+                      "type": [
+                        "string",
+                        "null"
+                      ],
+                      "description": "Likely intended domain, e.g. gmail.com for gmial.com."
+                    },
+                    "parkedDomain": {
+                      "type": [
+                        "boolean",
+                        "null"
+                      ],
+                      "description": "Parking name servers, typo-catcher MX or a for-sale / parked homepage. null = not checked."
+                    }
+                  },
+                  "description": "Flags"
+                },
+                "suggestion": {
+                  "type": [
+                    "string",
+                    "null"
+                  ],
+                  "description": "Corrected full address when a typo is suspected."
+                },
+                "domainHealth": {
+                  "type": "object",
+                  "properties": {
+                    "spf": {
+                      "type": [
+                        "string",
+                        "null"
+                      ],
+                      "description": "SPF 'all' qualifier: fail, softfail, neutral, pass, no_all; null = no SPF record."
+                    },
+                    "dmarc": {
+                      "type": [
+                        "string",
+                        "null"
+                      ],
+                      "description": "DMARC policy: none, quarantine, reject; null = no DMARC record."
+                    },
+                    "createdAt": {
+                      "type": [
+                        "string",
+                        "null"
+                      ],
+                      "description": "Domain registration date (RDAP)."
+                    },
+                    "ageDays": {
+                      "type": [
+                        "integer",
+                        "null"
+                      ]
+                    },
+                    "registrar": {
+                      "type": [
+                        "string",
+                        "null"
+                      ]
+                    },
+                    "parkedEvidence": {
+                      "type": [
+                        "string",
+                        "null"
+                      ]
+                    }
+                  },
+                  "description": "Domain health"
+                },
+                "catchAll": {
+                  "type": "string",
+                  "enum": [
+                    "not_checked"
+                  ],
+                  "description": "Catch-all detection needs SMTP; always not_checked in this version."
+                },
+                "mailbox": {
+                  "type": "string",
+                  "enum": [
+                    "not_checked"
+                  ],
+                  "description": "The individual mailbox is never probed over SMTP; always not_checked."
+                },
+                "errors": {
+                  "type": "array",
+                  "items": {
+                    "type": "string"
+                  },
+                  "description": "Non-fatal lookup errors (e.g. rdap: no public RDAP server for .xyz)."
+                },
+                "ms": {
+                  "type": "integer",
+                  "description": "Time spent on this address in milliseconds."
+                },
+                "checkedAt": {
+                  "type": "string",
+                  "format": "date-time",
+                  "description": "Checked at"
+                }
+              }
+            }
+          }
+        }
+      },
+      "annotations": {
+        "title": "Mailvett",
+        "readOnlyHint": true,
+        "destructiveHint": false,
+        "idempotentHint": false,
+        "openWorldHint": true
+      }
+    }
+  ],
+  "localTools": [
+    {
+      "name": "validate_emails",
+      "title": "Mailvett: validate email addresses",
+      "description": "Mailvett. Email validation / verification: give email addresses and get, per address, a VERDICT (deliverable_likely, risky, undeliverable, invalid), a 0-100 score and machine-readable reasons (e.g. mx_present, disposable_domain, role_account:sales, possible_typo_of:gmail.com, domain_not_found), plus the syntax check, MX hosts and mail provider (Gmail, Google Workspace, Microsoft 365, Zoho, Proton...), disposable, role, free-provider, typo and parked-domain flags, a corrected address when a typo is likely (gmial.com -> gmail.com), domain age (RDAP), SPF and DMARC. Use it to check an address BEFORE sending an email, saving a contact or accepting a sign-up, and to clean lead lists. It does not probe the mailbox over SMTP: mailbox and catchAll are always \"not_checked\". Cost: $0.0008 per address checked; invalid syntax and duplicates are free. About 1 s per new domain."
+    }
+  ]
+}

market/mcp/html-to-pdf.tools.json

@@ -5,13 +5,13 @@
     "url": "https://mcp.apify.com/?tools=cybermax/html-to-pdf",
     "transport": "streamable-http",
     "auth": "Authorization: Bearer <APIFY_TOKEN>",
-    "local": null
+    "local": "mcp/html-to-pdf-mcp"
   },
   "tools": [
     {
       "name": "cybermax--html-to-pdf",
       "title": "Printwren: HTML to PDF & URL to PDF API",
-      "description": "HTML to PDF and URL to PDF API with headless Chrome: A4/Letter/Legal, landscape, margins, header/footer with page numbers, print CSS. Returns a public PDF link, page count and size. Schedule it to archive web pages as PDFs. $0.003 per PDF, failures free. Pricing: PDF generated $0.003 (billed to the caller's Apify account).",
+      "description": "Printwren. HTML to PDF and URL to PDF with headless Chrome: give web page URLs and/or one complete raw HTML document (invoice, report, receipt, ticket) and it returns, per document: pdfUrl (public link that opens without a token), pages (page count), bytes, page title, finalUrl after redirects and httpStatus, plus the paper size used; failed documents come back in \"failed\" with the reason. Options: A4/Letter/Legal/A3/A5, landscape, margins, page numbers, HTML header/footer templates, print or screen CSS, wait for a selector, PDF bookmarks from headings, tagged PDF, optional base64 bytes. Cookie banners are dismissed automatically. Use it to render an invoice or report you just generated as a PDF, archive a web page as a dated PDF, or attach a printable copy of an article to an email or ticket. Not for screenshots (use SnapScout screenshot_website), Word/Excel/PowerPoint files (use Paperkiln office_to_pdf), or pages behind a login or CAPTCHA. Cost: $0.003 per PDF generated; failed pages are free. Pricing: PDF generated $0.003 (billed to the caller's Apify account).",
       "inputSchema": {
         "type": "object",
         "properties": {
@@ -295,5 +295,11 @@
       }
     }
   ],
-  "localTools": []
+  "localTools": [
+    {
+      "name": "html_to_pdf",
+      "title": "Printwren: convert web pages or raw HTML to PDF",
+      "description": "Printwren. HTML to PDF and URL to PDF with headless Chrome: give web page URLs and/or one complete raw HTML document (invoice, report, receipt, ticket) and it returns, per document: pdfUrl (public link that opens without a token), pages (page count), bytes, page title, finalUrl after redirects and httpStatus, plus the paper size used; failed documents come back in \"failed\" with the reason. Options: A4/Letter/Legal/A3/A5, landscape, margins, page numbers, HTML header/footer templates, print or screen CSS, wait for a selector, PDF bookmarks from headings, tagged PDF, optional base64 bytes. Cookie banners are dismissed automatically. Use it to render an invoice or report you just generated as a PDF, archive a web page as a dated PDF, or attach a printable copy of an article to an email or ticket. Not for screenshots (use SnapScout screenshot_website), Word/Excel/PowerPoint files (use Paperkiln office_to_pdf), or pages behind a login or CAPTCHA. Cost: $0.003 per PDF generated; failed pages are free."
+    }
+  ]
 }

market/mcp/office-to-pdf.tools.json

@@ -5,13 +5,13 @@
     "url": "https://mcp.apify.com/?tools=cybermax/office-to-pdf",
     "transport": "streamable-http",
     "auth": "Authorization: Bearer <APIFY_TOKEN>",
-    "local": null
+    "local": "mcp/office-to-pdf-mcp"
   },
   "tools": [
     {
       "name": "cybermax--office-to-pdf",
       "title": "Paperkiln: Word, Excel & PowerPoint to PDF Converter API",
-      "description": "Convert DOCX, XLSX, PPTX, ODT, RTF, CSV and legacy DOC/XLS/PPT files to PDF (or PDF/A) in bulk with headless LibreOffice and Office-compatible fonts. Give URLs or share links (Dropbox, Google Drive, OneDrive); get hosted PDFs with page count. $0.005 per file; failures free. Pricing: Document converted $0.005 (billed to the caller's Apify account).",
+      "description": "Paperkiln. Converts office documents to PDF (or archival PDF/A-2b) with headless LibreOffice and Office-compatible fonts: .docx .doc .xlsx .xls .pptx .ppt .odt .ods .odp .rtf .csv .txt .vsdx by URL, including public Dropbox, Google Drive, OneDrive/SharePoint and Google Docs/Sheets/Slides share links. Returns, per file: pdfUrl (shareable link that works without a token), pages (page count), pdfBytes, inputFormat, fileName, whether it is PDF/A and conversion time; failed files come back in \"failed\" with the reason. Use it to turn a generated DOCX or XLSX report into a PDF you can send, batch-convert a folder of contracts to PDF/A for archiving, or get a page count before an e-signature step. For web pages or raw HTML use Printwren html_to_pdf instead (it renders in a real browser). Not for PDF to Word, password-protected files or running macros. Cost: $0.005 per file converted; failed downloads, unsupported files and conversion errors are free. Pricing: Document converted $0.005 (billed to the caller's Apify account).",
       "inputSchema": {
         "type": "object",
         "properties": {
@@ -199,5 +199,11 @@
       }
     }
   ],
-  "localTools": []
+  "localTools": [
+    {
+      "name": "office_to_pdf",
+      "title": "Paperkiln: convert Word, Excel & PowerPoint files to PDF",
+      "description": "Paperkiln. Converts office documents to PDF (or archival PDF/A-2b) with headless LibreOffice and Office-compatible fonts: .docx .doc .xlsx .xls .pptx .ppt .odt .ods .odp .rtf .csv .txt .vsdx by URL, including public Dropbox, Google Drive, OneDrive/SharePoint and Google Docs/Sheets/Slides share links. Returns, per file: pdfUrl (shareable link that works without a token), pages (page count), pdfBytes, inputFormat, fileName, whether it is PDF/A and conversion time; failed files come back in \"failed\" with the reason. Use it to turn a generated DOCX or XLSX report into a PDF you can send, batch-convert a folder of contracts to PDF/A for archiving, or get a page count before an e-signature step. For web pages or raw HTML use Printwren html_to_pdf instead (it renders in a real browser). Not for PDF to Word, password-protected files or running macros. Cost: $0.005 per file converted; failed downloads, unsupported files and conversion errors are free."
+    }
+  ]
 }

market/mcp/rss-feed-finder.tools.json

@@ -5,13 +5,13 @@
     "url": "https://mcp.apify.com/?tools=cybermax/rss-feed-finder",
     "transport": "streamable-http",
     "auth": "Authorization: Bearer <APIFY_TOKEN>",
-    "local": null
+    "local": "mcp/rss-feed-finder-mcp"
   },
   "tools": [
     {
       "name": "cybermax--rss-feed-finder",
       "title": "Feedpeck: RSS Feed Finder & Reader: Discover Feeds by Website",
-      "description": "Find every RSS, Atom and JSON feed a website publishes (link tags, page links, common paths, platform rules for Reddit, Medium, YouTube, GitHub), validated, with title, type, last post date and the latest items. Monitor mode returns only new posts. $0.002 per feed found. Pricing: Feed found $0.002; Post returned $0.0003 (billed to the caller's Apify account).",
+      "description": "Feedpeck. RSS feed finder and reader: give websites (domains, blog sections like stripe.com/blog, subreddits, Medium profiles, YouTube channels, GitHub repos or feed URLs) and it discovers every RSS, Atom and JSON feed they publish (link tags, page links, common paths, platform rules), validates each, and returns per feed: feedUrl, feedTitle, feedType, discoveredVia, lastItemAt (date of the newest post), itemCount and latestItems (title, url, published, summary, image). With output=\"items\" it returns one row per post instead; onlyNew=true returns only posts not seen on an earlier call. Sites with no feed are listed in sitesWithoutFeed. Use it to get the feed URL for a blog, check whether a company or competitor is still publishing (lastItemAt), or pull the latest posts from 20 sites for a digest. Feeds often carry only a summary: for full article text pass the post URLs to Inkpluck extract_article. Cost: $0.002 per feed found (latest posts included), or $0.0003 per post with output=\"items\" or onlyNew; sites without a feed are free. Pricing: Feed found $0.002; Post returned $0.0003 (billed to the caller's Apify account).",
       "inputSchema": {
         "type": "object",
         "properties": {
@@ -247,5 +247,11 @@
       }
     }
   ],
-  "localTools": []
+  "localTools": [
+    {
+      "name": "find_rss_feeds",
+      "title": "Feedpeck: find a website's RSS/Atom feeds and read its latest posts",
+      "description": "Feedpeck. RSS feed finder and reader: give websites (domains, blog sections like stripe.com/blog, subreddits, Medium profiles, YouTube channels, GitHub repos or feed URLs) and it discovers every RSS, Atom and JSON feed they publish (link tags, page links, common paths, platform rules), validates each, and returns per feed: feedUrl, feedTitle, feedType, discoveredVia, lastItemAt (date of the newest post), itemCount and latestItems (title, url, published, summary, image). With output=\"items\" it returns one row per post instead; onlyNew=true returns only posts not seen on an earlier call. Sites with no feed are listed in sitesWithoutFeed. Use it to get the feed URL for a blog, check whether a company or competitor is still publishing (lastItemAt), or pull the latest posts from 20 sites for a digest. Feeds often carry only a summary: for full article text pass the post URLs to Inkpluck extract_article. Cost: $0.002 per feed found (latest posts included), or $0.0003 per post with output=\"items\" or onlyNew; sites without a feed are free."
+    }
+  ]
 }

market/openapi/datasets.json

@@ -0,0 +1,578 @@
+{
+  "openapi": "3.1.0",
+  "info": {
+    "title": "CyberMax open datasets API (Hugging Face dataset viewer)",
+    "version": "1.0.0",
+    "summary": "Query 21 free CyberMax datasets without downloading them.",
+    "description": "The public Hugging Face dataset viewer API, documented for the CyberMax datasets only. No key, free. Full-text search and filters use an index built on first use. Catalog: https://shaw276-cybermax.static.hf.space/market/catalog.json",
+    "contact": {
+      "name": "CyberMax",
+      "url": "https://shaw276-cybermax.static.hf.space/market/llms.txt"
+    }
+  },
+  "externalDocs": {
+    "description": "Hugging Face dataset viewer docs",
+    "url": "https://huggingface.co/docs/dataset-viewer"
+  },
+  "servers": [
+    {
+      "url": "https://datasets-server.huggingface.co"
+    }
+  ],
+  "security": [],
+  "tags": [
+    {
+      "name": "datasets",
+      "description": "CyberMax open datasets"
+    }
+  ],
+  "paths": {
+    "/splits": {
+      "get": {
+        "operationId": "listSplits",
+        "tags": [
+          "datasets"
+        ],
+        "summary": "Configs and splits of a dataset",
+        "parameters": [
+          {
+            "name": "dataset",
+            "in": "query",
+            "required": true,
+            "description": "CyberMax dataset id.",
+            "schema": {
+              "type": "string",
+              "enum": [
+                "shaw276/open-domain-ranks",
+                "shaw276/saas-email-security-snapshot",
+                "shaw276/tech-hiring-snapshot",
+                "shaw276/app-complaint-stats",
+                "shaw276/sec-form4-insider-transactions",
+                "shaw276/kevscope-cve-kev-epss-daily",
+                "shaw276/tickerbell-sec-8k-events",
+                "shaw276/us-zip-codes-demographics",
+                "shaw276/curvewire-us-macro-panel",
+                "shaw276/trialtide-clinicaltrials-registry",
+                "shaw276/fipsbook-us-census-geographies",
+                "shaw276/toolroll-mcp-registry-snapshot",
+                "shaw276/tallyfold-sec-xbrl-financials",
+                "shaw276/pingroll-mcp-server-uptime",
+                "shaw276/riskroll-sec-10k-10q-sections",
+                "shaw276/policyloom-us-ai-policy-tracker",
+                "shaw276/swellmeter-google-trending-searches",
+                "shaw276/hireheat-career-site-jobs-weekly",
+                "shaw276/smartthings-review-mood-sept-2026",
+                "shaw276/superintelligence-search-questions",
+                "shaw276/superintelligence-ban-bill-coverage"
+              ]
+            }
+          }
+        ],
+        "responses": {
+          "200": {
+            "description": "Splits.",
+            "content": {
+              "application/json": {
+                "schema": {
+                  "type": "object",
+                  "properties": {
+                    "splits": {
+                      "type": "array",
+                      "items": {
+                        "type": "object",
+                        "properties": {
+                          "dataset": {
+                            "type": "string"
+                          },
+                          "config": {
+                            "type": "string"
+                          },
+                          "split": {
+                            "type": "string"
+                          }
+                        }
+                      }
+                    }
+                  }
+                }
+              }
+            }
+          },
+          "404": {
+            "description": "Unknown dataset/config/split."
+          },
+          "500": {
+            "description": "Index still loading or viewer error; retry later."
+          }
+        }
+      }
+    },
+    "/first-rows": {
+      "get": {
+        "operationId": "firstRows",
+        "tags": [
+          "datasets"
+        ],
+        "summary": "Column types and the first rows",
+        "parameters": [
+          {
+            "name": "dataset",
+            "in": "query",
+            "required": true,
+            "description": "CyberMax dataset id.",
+            "schema": {
+              "type": "string",
+              "enum": [
+                "shaw276/open-domain-ranks",
+                "shaw276/saas-email-security-snapshot",
+                "shaw276/tech-hiring-snapshot",
+                "shaw276/app-complaint-stats",
+                "shaw276/sec-form4-insider-transactions",
+                "shaw276/kevscope-cve-kev-epss-daily",
+                "shaw276/tickerbell-sec-8k-events",
+                "shaw276/us-zip-codes-demographics",
+                "shaw276/curvewire-us-macro-panel",
+                "shaw276/trialtide-clinicaltrials-registry",
+                "shaw276/fipsbook-us-census-geographies",
+                "shaw276/toolroll-mcp-registry-snapshot",
+                "shaw276/tallyfold-sec-xbrl-financials",
+                "shaw276/pingroll-mcp-server-uptime",
+                "shaw276/riskroll-sec-10k-10q-sections",
+                "shaw276/policyloom-us-ai-policy-tracker",
+                "shaw276/swellmeter-google-trending-searches",
+                "shaw276/hireheat-career-site-jobs-weekly",
+                "shaw276/smartthings-review-mood-sept-2026",
+                "shaw276/superintelligence-search-questions",
+                "shaw276/superintelligence-ban-bill-coverage"
+              ]
+            }
+          },
+          {
+            "name": "config",
+            "in": "query",
+            "required": true,
+            "description": "Config (subset) name, from /splits.",
+            "schema": {
+              "type": "string"
+            }
+          },
+          {
+            "name": "split",
+            "in": "query",
+            "required": true,
+            "description": "Split name, from /splits (usually train).",
+            "schema": {
+              "type": "string"
+            }
+          }
+        ],
+        "responses": {
+          "200": {
+            "description": "Rows.",
+            "content": {
+              "application/json": {
+                "schema": {
+                  "$ref": "#/components/schemas/Rows"
+                }
+              }
+            }
+          },
+          "404": {
+            "description": "Unknown dataset/config/split."
+          },
+          "500": {
+            "description": "Index still loading or viewer error; retry later."
+          }
+        }
+      }
+    },
+    "/rows": {
+      "get": {
+        "operationId": "getRows",
+        "tags": [
+          "datasets"
+        ],
+        "summary": "Page through rows",
+        "parameters": [
+          {
+            "name": "dataset",
+            "in": "query",
+            "required": true,
+            "description": "CyberMax dataset id.",
+            "schema": {
+              "type": "string",
+              "enum": [
+                "shaw276/open-domain-ranks",
+                "shaw276/saas-email-security-snapshot",
+                "shaw276/tech-hiring-snapshot",
+                "shaw276/app-complaint-stats",
+                "shaw276/sec-form4-insider-transactions",
+                "shaw276/kevscope-cve-kev-epss-daily",
+                "shaw276/tickerbell-sec-8k-events",
+                "shaw276/us-zip-codes-demographics",
+                "shaw276/curvewire-us-macro-panel",
+                "shaw276/trialtide-clinicaltrials-registry",
+                "shaw276/fipsbook-us-census-geographies",
+                "shaw276/toolroll-mcp-registry-snapshot",
+                "shaw276/tallyfold-sec-xbrl-financials",
+                "shaw276/pingroll-mcp-server-uptime",
+                "shaw276/riskroll-sec-10k-10q-sections",
+                "shaw276/policyloom-us-ai-policy-tracker",
+                "shaw276/swellmeter-google-trending-searches",
+                "shaw276/hireheat-career-site-jobs-weekly",
+                "shaw276/smartthings-review-mood-sept-2026",
+                "shaw276/superintelligence-search-questions",
+                "shaw276/superintelligence-ban-bill-coverage"
+              ]
+            }
+          },
+          {
+            "name": "config",
+            "in": "query",
+            "required": true,
+            "description": "Config (subset) name, from /splits.",
+            "schema": {
+              "type": "string"
+            }
+          },
+          {
+            "name": "split",
+            "in": "query",
+            "required": true,
+            "description": "Split name, from /splits (usually train).",
+            "schema": {
+              "type": "string"
+            }
+          },
+          {
+            "name": "offset",
+            "in": "query",
+            "required": false,
+            "description": "First row (0-based).",
+            "schema": {
+              "type": "integer",
+              "minimum": 0,
+              "default": 0
+            }
+          },
+          {
+            "name": "length",
+            "in": "query",
+            "required": false,
+            "description": "Rows to return (max 100).",
+            "schema": {
+              "type": "integer",
+              "minimum": 1,
+              "maximum": 100,
+              "default": 100
+            }
+          }
+        ],
+        "responses": {
+          "200": {
+            "description": "Rows.",
+            "content": {
+              "application/json": {
+                "schema": {
+                  "$ref": "#/components/schemas/Rows"
+                }
+              }
+            }
+          },
+          "404": {
+            "description": "Unknown dataset/config/split."
+          },
+          "500": {
+            "description": "Index still loading or viewer error; retry later."
+          }
+        }
+      }
+    },
+    "/search": {
+      "get": {
+        "operationId": "searchRows",
+        "tags": [
+          "datasets"
+        ],
+        "summary": "Full-text search over string columns",
+        "parameters": [
+          {
+            "name": "dataset",
+            "in": "query",
+            "required": true,
+            "description": "CyberMax dataset id.",
+            "schema": {
+              "type": "string",
+              "enum": [
+                "shaw276/open-domain-ranks",
+                "shaw276/saas-email-security-snapshot",
+                "shaw276/tech-hiring-snapshot",
+                "shaw276/app-complaint-stats",
+                "shaw276/sec-form4-insider-transactions",
+                "shaw276/kevscope-cve-kev-epss-daily",
+                "shaw276/tickerbell-sec-8k-events",
+                "shaw276/us-zip-codes-demographics",
+                "shaw276/curvewire-us-macro-panel",
+                "shaw276/trialtide-clinicaltrials-registry",
+                "shaw276/fipsbook-us-census-geographies",
+                "shaw276/toolroll-mcp-registry-snapshot",
+                "shaw276/tallyfold-sec-xbrl-financials",
+                "shaw276/pingroll-mcp-server-uptime",
+                "shaw276/riskroll-sec-10k-10q-sections",
+                "shaw276/policyloom-us-ai-policy-tracker",
+                "shaw276/swellmeter-google-trending-searches",
+                "shaw276/hireheat-career-site-jobs-weekly",
+                "shaw276/smartthings-review-mood-sept-2026",
+                "shaw276/superintelligence-search-questions",
+                "shaw276/superintelligence-ban-bill-coverage"
+              ]
+            }
+          },
+          {
+            "name": "config",
+            "in": "query",
+            "required": true,
+            "description": "Config (subset) name, from /splits.",
+            "schema": {
+              "type": "string"
+            }
+          },
+          {
+            "name": "split",
+            "in": "query",
+            "required": true,
+            "description": "Split name, from /splits (usually train).",
+            "schema": {
+              "type": "string"
+            }
+          },
+          {
+            "name": "query",
+            "in": "query",
+            "required": true,
+            "description": "Search text.",
+            "schema": {
+              "type": "string"
+            }
+          },
+          {
+            "name": "offset",
+            "in": "query",
+            "required": false,
+            "description": "First row (0-based).",
+            "schema": {
+              "type": "integer",
+              "minimum": 0,
+              "default": 0
+            }
+          },
+          {
+            "name": "length",
+            "in": "query",
+            "required": false,
+            "description": "Rows to return (max 100).",
+            "schema": {
+              "type": "integer",
+              "minimum": 1,
+              "maximum": 100,
+              "default": 100
+            }
+          }
+        ],
+        "responses": {
+          "200": {
+            "description": "Rows.",
+            "content": {
+              "application/json": {
+                "schema": {
+                  "$ref": "#/components/schemas/Rows"
+                }
+              }
+            }
+          },
+          "404": {
+            "description": "Unknown dataset/config/split."
+          },
+          "500": {
+            "description": "Index still loading or viewer error; retry later."
+          }
+        }
+      }
+    },
+    "/filter": {
+      "get": {
+        "operationId": "filterRows",
+        "tags": [
+          "datasets"
+        ],
+        "summary": "SQL-style filter and sort (DuckDB syntax)",
+        "parameters": [
+          {
+            "name": "dataset",
+            "in": "query",
+            "required": true,
+            "description": "CyberMax dataset id.",
+            "schema": {
+              "type": "string",
+              "enum": [
+                "shaw276/open-domain-ranks",
+                "shaw276/saas-email-security-snapshot",
+                "shaw276/tech-hiring-snapshot",
+                "shaw276/app-complaint-stats",
+                "shaw276/sec-form4-insider-transactions",
+                "shaw276/kevscope-cve-kev-epss-daily",
+                "shaw276/tickerbell-sec-8k-events",
+                "shaw276/us-zip-codes-demographics",
+                "shaw276/curvewire-us-macro-panel",
+                "shaw276/trialtide-clinicaltrials-registry",
+                "shaw276/fipsbook-us-census-geographies",
+                "shaw276/toolroll-mcp-registry-snapshot",
+                "shaw276/tallyfold-sec-xbrl-financials",
+                "shaw276/pingroll-mcp-server-uptime",
+                "shaw276/riskroll-sec-10k-10q-sections",
+                "shaw276/policyloom-us-ai-policy-tracker",
+                "shaw276/swellmeter-google-trending-searches",
+                "shaw276/hireheat-career-site-jobs-weekly",
+                "shaw276/smartthings-review-mood-sept-2026",
+                "shaw276/superintelligence-search-questions",
+                "shaw276/superintelligence-ban-bill-coverage"
+              ]
+            }
+          },
+          {
+            "name": "config",
+            "in": "query",
+            "required": true,
+            "description": "Config (subset) name, from /splits.",
+            "schema": {
+              "type": "string"
+            }
+          },
+          {
+            "name": "split",
+            "in": "query",
+            "required": true,
+            "description": "Split name, from /splits (usually train).",
+            "schema": {
+              "type": "string"
+            }
+          },
+          {
+            "name": "where",
+            "in": "query",
+            "required": false,
+            "description": "Condition, e.g. \"state\"='NJ' AND \"population\">50000.",
+            "schema": {
+              "type": "string"
+            }
+          },
+          {
+            "name": "orderby",
+            "in": "query",
+            "required": false,
+            "description": "Sort, e.g. \"population\" DESC.",
+            "schema": {
+              "type": "string"
+            }
+          },
+          {
+            "name": "offset",
+            "in": "query",
+            "required": false,
+            "description": "First row (0-based).",
+            "schema": {
+              "type": "integer",
+              "minimum": 0,
+              "default": 0
+            }
+          },
+          {
+            "name": "length",
+            "in": "query",
+            "required": false,
+            "description": "Rows to return (max 100).",
+            "schema": {
+              "type": "integer",
+              "minimum": 1,
+              "maximum": 100,
+              "default": 100
+            }
+          }
+        ],
+        "responses": {
+          "200": {
+            "description": "Rows.",
+            "content": {
+              "application/json": {
+                "schema": {
+                  "$ref": "#/components/schemas/Rows"
+                }
+              }
+            }
+          },
+          "404": {
+            "description": "Unknown dataset/config/split."
+          },
+          "500": {
+            "description": "Index still loading or viewer error; retry later."
+          }
+        }
+      }
+    }
+  },
+  "components": {
+    "schemas": {
+      "Rows": {
+        "type": "object",
+        "properties": {
+          "features": {
+            "type": "array",
+            "items": {
+              "type": "object",
+              "properties": {
+                "feature_idx": {
+                  "type": "integer"
+                },
+                "name": {
+                  "type": "string"
+                },
+                "type": {
+                  "type": "object"
+                }
+              }
+            }
+          },
+          "rows": {
+            "type": "array",
+            "items": {
+              "type": "object",
+              "properties": {
+                "row_idx": {
+                  "type": "integer"
+                },
+                "row": {
+                  "type": "object"
+                },
+                "truncated_cells": {
+                  "type": "array",
+                  "items": {
+                    "type": "string"
+                  }
+                }
+              }
+            }
+          },
+          "num_rows_total": {
+            "type": "integer"
+          },
+          "num_rows_per_page": {
+            "type": "integer"
+          },
+          "partial": {
+            "type": "boolean"
+          }
+        }
+      }
+    }
+  }
+}

market/schemas/catalog.schema.json

@@ -49,7 +49,7 @@
       "type": "object",
       "required": ["id", "kind", "status", "brand", "title", "summary", "description", "capabilities", "interfaces", "outputs", "api", "docs", "pricing", "integration", "usageNotes", "publisher"],
       "properties": {
-        "id": { "type": "string", "pattern": "^(cybermax/[a-z0-9-]+|hf:[A-Za-z0-9-]+/[A-Za-z0-9._-]+)$" },
+        "id": { "type": "string", "pattern": "^(cybermax/[a-z0-9-]+|hf:[A-Za-z0-9-]+/[A-Za-z0-9._-]+|web:[a-z0-9-]+)$" },
         "kind": { "enum": ["api_tool", "dataset", "web_app"] },
         "status": { "enum": ["live", "coming_soon"] },
         "brand": { "type": "string", "minLength": 1 },
@@ -99,9 +99,10 @@
             { "type": "null" },
             {
               "type": "object",
-              "required": ["hosted"],
+              "anyOf": [{ "required": ["hosted"], "properties": { "hosted": { "type": "object" } } }, { "required": ["local"], "properties": { "local": { "type": "object" } } }],
               "properties": {
-                "hosted": { "type": "object", "required": ["url", "transport", "toolName"], "properties": { "transport": { "const": "streamable-http" }, "toolName": { "type": "string", "pattern": "^[A-Za-z0-9_.-]{1,128}$" } } }
+                "hosted": { "oneOf": [{ "type": "null" }, { "type": "object", "required": ["url", "transport", "toolName"], "properties": { "transport": { "const": "streamable-http" }, "toolName": { "type": "string", "pattern": "^[A-Za-z0-9_.-]{1,128}$" } } }] },
+                "local": { "oneOf": [{ "type": "null" }, { "type": "object", "required": ["path", "tools"] }] }
               }
             }
           ]

market/tests/picker-tasks.json

@@ -1,17 +1,76 @@
 {
   "description": "Sample agent tasks for the tool-picker test (market/TESTS.md). expected = catalog id, or null when no CyberMax capability fits.",
   "tasks": [
-    { "id": 1, "task": "Does figma.com enforce DMARC, and when does its TLS certificate expire?", "expected": "cybermax/domain-intel" },
-    { "id": 2, "task": "Find the support email address and the LinkedIn page for linear.app and vercel.com.", "expected": "cybermax/website-contacts" },
-    { "id": 3, "task": "Is US interest in 'standing desk' rising or falling over the past 12 months?", "expected": "cybermax/google-trends" },
-    { "id": 4, "task": "List remote data engineer jobs posted in the last 7 days, with apply links.", "expected": "cybermax/career-site-jobs" },
-    { "id": 5, "task": "How many engineers is Datadog hiring right now, and is hiring speeding up?", "expected": "cybermax/hiring-signals" },
-    { "id": 6, "task": "Have any directors or officers of NVDA bought shares on the open market in the last two weeks?", "expected": "cybermax/sec-insider-tracker" },
-    { "id": 7, "task": "What are Duolingo users complaining about in their recent 1- and 2-star reviews?", "expected": "cybermax/app-reviews" },
-    { "id": 8, "task": "Give me 200 long-tail keyword ideas and questions people search around 'cold brew coffee'.", "expected": "cybermax/keyword-ideas" },
-    { "id": 9, "task": "Sort these 500 prospect domains by how big and established each website is.", "expected": "cybermax/domain-authority" },
-    { "id": 10, "task": "Show me what stripe.com/pricing looks like on a phone screen.", "expected": "cybermax/website-screenshot" },
-    { "id": 11, "task": "I need a free downloadable table of every study registered on ClinicalTrials.gov.", "expected": "hf:shaw276/trialtide-clinicaltrials-registry" },
-    { "id": 12, "task": "Translate this paragraph into French.", "expected": null }
+    {
+      "id": 1,
+      "task": "Does figma.com enforce DMARC, and when does its TLS certificate expire?",
+      "expected": "cybermax/domain-intel"
+    },
+    {
+      "id": 2,
+      "task": "Find the support email address and the LinkedIn page for linear.app and vercel.com.",
+      "expected": "cybermax/website-contacts"
+    },
+    {
+      "id": 3,
+      "task": "Is US interest in 'standing desk' rising or falling over the past 12 months?",
+      "expected": "cybermax/google-trends"
+    },
+    {
+      "id": 4,
+      "task": "List remote data engineer jobs posted in the last 7 days, with apply links.",
+      "expected": "cybermax/career-site-jobs"
+    },
+    {
+      "id": 5,
+      "task": "How many engineers is Datadog hiring right now, and is hiring speeding up?",
+      "expected": "cybermax/hiring-signals"
+    },
+    {
+      "id": 6,
+      "task": "Have any directors or officers of NVDA bought shares on the open market in the last two weeks?",
+      "expected": "cybermax/sec-insider-tracker"
+    },
+    {
+      "id": 7,
+      "task": "What are Duolingo users complaining about in their recent 1- and 2-star reviews?",
+      "expected": "cybermax/app-reviews"
+    },
+    {
+      "id": 8,
+      "task": "Give me 200 long-tail keyword ideas and questions people search around 'cold brew coffee'.",
+      "expected": "cybermax/keyword-ideas"
+    },
+    {
+      "id": 9,
+      "task": "Sort these 500 prospect domains by how big and established each website is.",
+      "expected": "cybermax/domain-authority"
+    },
+    {
+      "id": 10,
+      "task": "Show me what stripe.com/pricing looks like on a phone screen.",
+      "expected": "cybermax/website-screenshot"
+    },
+    {
+      "id": 11,
+      "task": "I need a free downloadable table of every study registered on ClinicalTrials.gov.",
+      "expected": "hf:shaw276/trialtide-clinicaltrials-registry"
+    },
+    {
+      "id": 12,
+      "task": "Translate this paragraph into French.",
+      "expected": null
+    },
+    {
+      "id": 13,
+      "task": "Before emailing these 300 sign-ups, flag disposable and mistyped addresses.",
+      "expected": "cybermax/email-validate",
+      "note": "coming soon until Mailvett is public; pickers may choose coming-soon entries only when nothing live fits"
+    },
+    {
+      "id": 14,
+      "task": "My user wants a simple page to work out what hourly rate a $85,000 salary equals.",
+      "expected": "web:salary-to-hourly-calculator"
+    }
   ]
 }

market/tools/ai-extract.md

@@ -0,0 +1,50 @@
+# Fieldwright: AI Web Data Extractor (Bring Your Own LLM Key)
+
+> AI web scraper that turns any page into typed JSON fields you name, using YOUR OpenAI, Claude, Gemini or OpenRouter key (you pay the model at cost).
+
+**Status: coming soon, not callable yet.**
+
+AI web scraper that turns any page into typed JSON fields you name, using YOUR OpenAI, Claude, Gemini or OpenRouter key (you pay the model at cost). List mode, change monitoring, JSON-LD. $0.001/page; failures free.
+
+- ID: `cybermax/ai-extract` · MCP tool: `cybermax--ai-extract` at https://mcp.apify.com/?tools=cybermax/ai-extract
+- HTTP: `POST https://api.apify.com/v2/acts/cybermax~ai-extract/run-sync-get-dataset-items` (header `Authorization: Bearer <APIFY_TOKEN>`, body = input JSON, returns a JSON array)
+- Price: $0.001 per page extracted with your llm + $0.0003 per page parsed without llm (pay per event, billed by Apify to the caller's account)
+- Store page: https://apify.com/cybermax/ai-extract
+- Input fields: `urls` (array), `fields` (array), `instructions` (string), `extractList` (boolean), `jsonSchema` (object), `llmProvider` (string), `llmApiKey` (string), `llmModel` (string), `llmBaseUrl` (string), `maxOutputTokens` (integer), `maxPageChars` (integer), `onlyChanged` (boolean), `monitorName` (string), `includeMarkdown` (boolean), `includeMetadata` (boolean), `htmlDocuments` (array), `maxPages` (integer), `maxConcurrency` (integer), `timeoutSecs` (integer), `retryBlockedWithProxy` (boolean), `proxyConfiguration` (object)
+- Output fields: `url`, `finalUrl`, `ok`, `mode`, `title`, `data`, `missingFields`, `itemIndex`, `itemsFound`, `changed`, `llm`, `description`, `language`, `canonicalUrl`, `openGraph`, `jsonLd`, `emails`, `phones`, `markdown`, `pageChars`, `truncated`, `httpStatus`, `fetchedVia`, `ms`, `error`, `extractedAt`
+- One record per input item; invalid or unresolvable inputs are returned with an error field and are not charged unless the pricing says otherwise.
+- On the hosted MCP server the tool is named "cybermax--ai-extract" and takes the same input object (plus optional waitSecs).
+
+Example input:
+
+```json
+{"urls":["https://books.toscrape.com/catalogue/category/books/poetry_23/index.html"],"extractList":true,"fields":[{"name":"title","type":"string"},{"name":"price","type":"number","description":"Price in GBP, number only"},{"name":"product_url","type":"url"},{"name":"in_stock","type":"boolean"}],"llmProvider":"openai","llmApiKey":"YOUR_OPENAI_KEY"}
+```
+
+Python:
+
+```python
+from apify_client import ApifyClient  # pip install apify-client
+client = ApifyClient("<APIFY_TOKEN>")
+run = client.actor("cybermax/ai-extract").call(run_input={"urls":["https://books.toscrape.com/catalogue/category/books/poetry_23/index.html"],"extractList":True,"fields":[{"name":"title","type":"string"},{"name":"price","type":"number","description":"Price in GBP, number only"},{"name":"product_url","type":"url"},{"name":"in_stock","type":"boolean"}],"llmProvider":"openai","llmApiKey":"YOUR_OPENAI_KEY"})
+for item in client.dataset(run["defaultDatasetId"]).iterate_items():
+    print(item)
+```
+
+MCP client config:
+
+```json
+{
+  "mcpServers": {
+    "cybermax-ai-extract": {
+      "type": "http",
+      "url": "https://mcp.apify.com/?tools=cybermax/ai-extract",
+      "headers": {
+        "Authorization": "Bearer <APIFY_TOKEN>"
+      }
+    }
+  }
+}
+```
+
+Catalog: https://shaw276-cybermax.static.hf.space/market/catalog.json · All tools: https://shaw276-cybermax.static.hf.space/market/llms.txt

market/tools/creative-commons-images.md

@@ -4,7 +4,11 @@
 
 **Status: coming soon, not callable yet.**
 
-Search free-to-use images by keyword across Openverse (900M+ CC images), NASA, The Met, Art Institute of Chicago, Cleveland Museum of Art and Wikimedia Commons. Every image comes with its license, commercial-use and modification flags, and a ready-to-paste attribution line. $0.001 per image.
+Commonkite. Free image search by keyword across Openverse (900M+ Creative Commons images from Flickr, Wikimedia and museums), NASA, The Met, Art Institute of Chicago, Cleveland Museum of Art and optionally Wikimedia Commons live search. Returns, per image: imageUrl, thumbnailUrl, width, height, title, creator, landingUrl, source, licenseName and licenseUrl, the flags commercialUse, modificationsAllowed and attributionRequired, and attributionText (a ready-to-paste credit line). License filter: "commercial" (default), "commercial-modify", "no-attribution" (CC0/public domain only) or "any". mode="random" returns one random image per term (placeholder/hero images). Use it to find a hero image for a blog post you are writing, illustrate slides with properly credited photos, or pick CC0 images that need no credit. Not for photos of specific products, logos, celebrities or today's news (rarely openly licensed); licenses are reported as the source states them. Cost: $0.001 per image returned; searches with no results are free.
+
+Use when: find a hero image for a blog post you are writing, illustrate slides with properly credited photos, or pick CC0 images that need no credit.
+
+Not for: Not for photos of specific products, logos, celebrities or today's news (rarely openly licensed); licenses are reported as the source states them.
 
 - ID: `cybermax/creative-commons-images` · MCP tool: `cybermax--creative-commons-images` at https://mcp.apify.com/?tools=cybermax/creative-commons-images
 - HTTP: `POST https://api.apify.com/v2/acts/cybermax~creative-commons-images/run-sync-get-dataset-items` (header `Authorization: Bearer <APIFY_TOKEN>`, body = input JSON, returns a JSON array)

market/tools/domain-verify.md

@@ -0,0 +1,55 @@
+# Bonafido: Business Domain Verification API
+
+> Checks whether a domain, URL or email domain is a real, operating, legitimate business website and (optionally) whether it belongs to the company you expect; returns a verdict, a 0-100 trust score, machine-readable risk flags with reasons, and quotable evidence.
+
+**Status: coming soon, not callable yet.**
+
+Bonafido. For each domain, URL or email (optionally with the company name you expect) returns one JSON record: verdict (verified | likely_legit | unverified | suspicious | not_found), a 0-100 trustScore, riskFlags (parked, new_domain_lt_30d/lt_90d, lookalike_of:<domain>, name_mismatch, not_official_domain, no_website, placeholder_page, tls_invalid, free_email_domain, redirects_offsite, no_mx, no_dmarc, no_contact_info) each with a reason, and evidence strings with their source (RDAP domain age, TLS, MX/SPF/DMARC, homepage, contact page/email/phone/address, social profiles, Wikidata official website, company-name match). Use it to decide whether a vendor, payee, lead or sign-up website is a real operating business and whether it belongs to the named company before paying, onboarding or emailing. Signals only, never a fraud finding. For raw tech-stack, DNS, security-header or WHOIS data use DomainDNA analyze_domains instead. Costs $0.006 per existing domain plus $0.004 when a company name is checked, billed to your Apify account; invalid and non-existent domains are free. About 1-3 s per domain, no browser.
+
+Use when: an agent must decide whether a vendor, supplier, payee, lead or partner website is a real operating business before acting; you need to confirm that a domain or email domain belongs to a named company (KYB-lite, onboarding, invoice or payee checks); you want to catch parked, for-sale, brand-new, look-alike or placeholder domains in a lead list or signup flow; you need evidence strings an LLM can quote to justify a trust decision.
+
+Not for: registry-grade KYB or sanctions screening (no company-registry, sanctions or beneficial-owner data); judging fraud or intent: the output is signals only, never a fraud finding; raw tech-stack, DNS, security-header or WHOIS data: use DomainDNA (cybermax/domain-intel) instead; finding a company's domain from its name alone: use Namewhere (cybermax/company-to-domain) instead; checking whether a mailbox exists: use an email verification tool.
+
+- ID: `cybermax/domain-verify` · MCP tool: `cybermax--domain-verify` at https://mcp.apify.com/?tools=cybermax/domain-verify
+- HTTP: `POST https://api.apify.com/v2/acts/cybermax~domain-verify/run-sync-get-dataset-items` (header `Authorization: Bearer <APIFY_TOKEN>`, body = input JSON, returns a JSON array)
+- Price: $0.006 per domain verified + $0.004 per company name checked (pay per event, billed by Apify to the caller's account)
+- Store page: https://apify.com/cybermax/domain-verify
+- Input fields: `items` (array, required), `country` (string), `checkContacts` (boolean), `maxContactPages` (integer), `useWikidata` (boolean), `timeoutSecs` (integer), `maxConcurrency` (integer), `monitor` (boolean), `monitorKey` (string)
+- Output fields: `input`, `domain`, `registrableDomain`, `companyName`, `country`, `verdict`, `trustScore`, `summary`, `riskFlags`, `riskDetails`, `evidence`, `resolves`, `reachable`, `httpStatus`, `finalUrl`, `redirectsToOtherDomain`, `redirectTarget`, `pageType`, `siteName`, `domainAgeDays`, `registeredAt`, `registrar`, `tlsValid`, `tlsDaysLeft`, `tlsIssuer`, `mx`, `mxProvider`, `spf`, `dmarcPolicy`, `contact`, `socials`, `nameMatch`, `lookalike`, `charged`, `errors`, `change`, `changes`, `checkedAt`, `ms`
+- Monitor mode: set "monitor": true on a scheduled run to get only records that changed since the last run.
+- One record per input item; invalid or unresolvable inputs are returned with an error field and are not charged unless the pricing says otherwise.
+- On the hosted MCP server the tool is named "cybermax--domain-verify" and takes the same input object (plus optional waitSecs).
+
+Example input:
+
+```json
+{"items":[{"domain":"stripe.com","companyName":"Stripe"},{"domain":"shopify.com","companyName":"Stripe"}]}
+```
+
+Python:
+
+```python
+from apify_client import ApifyClient  # pip install apify-client
+client = ApifyClient("<APIFY_TOKEN>")
+run = client.actor("cybermax/domain-verify").call(run_input={"items":[{"domain":"stripe.com","companyName":"Stripe"},{"domain":"shopify.com","companyName":"Stripe"}]})
+for item in client.dataset(run["defaultDatasetId"]).iterate_items():
+    print(item)
+```
+
+MCP client config:
+
+```json
+{
+  "mcpServers": {
+    "cybermax-domain-verify": {
+      "type": "http",
+      "url": "https://mcp.apify.com/?tools=cybermax/domain-verify",
+      "headers": {
+        "Authorization": "Bearer <APIFY_TOKEN>"
+      }
+    }
+  }
+}
+```
+
+Catalog: https://shaw276-cybermax.static.hf.space/market/catalog.json · All tools: https://shaw276-cybermax.static.hf.space/market/llms.txt

market/tools/email-validate.md

@@ -0,0 +1,54 @@
+# Mailvett: Email Validation API for AI Agents
+
+> Validate email addresses before sending or saving them: syntax, domain and MX / mail provider, disposable, role, free-provider and typo flags, domain age, SPF/DMARC and parked-domain signals, returned as a verdict (deliverable_likely, risky, undeliverable, invalid) with reason codes and a 0-100…
+
+**Status: coming soon, not callable yet.**
+
+Mailvett. Email validation / verification: give email addresses and get, per address, a VERDICT (deliverable_likely, risky, undeliverable, invalid), a 0-100 score and machine-readable reasons (e.g. mx_present, disposable_domain, role_account:sales, possible_typo_of:gmail.com, domain_not_found), plus the syntax check, MX hosts and mail provider (Gmail, Google Workspace, Microsoft 365, Zoho, Proton...), disposable, role, free-provider, typo and parked-domain flags, a corrected address when a typo is likely (gmial.com -> gmail.com), domain age (RDAP), SPF and DMARC. Use it to check an address BEFORE sending an email, saving a contact or accepting a sign-up, and to clean lead lists. It does not probe the mailbox over SMTP: mailbox and catchAll are always "not_checked". Cost: $0.0008 per address checked; invalid syntax and duplicates are free. About 1 s per new domain.
+
+Use when: you are about to send an email, save a contact or accept a sign-up and need to know whether the address can receive mail; cleaning a lead or CRM list: drop undeliverable and invalid addresses, flag disposable, role and free-provider ones; a user typed an address that might be misspelt (gmial.com, yahoo.con) and you want the corrected suggestion; screening sign-ups for throwaway (disposable) addresses or brand-new / parked domains.
+
+Not for: Confirming that one specific mailbox exists: no SMTP probe is made, so mailbox and catchAll are always not_checked; Finding someone's email address from a name or company (use an email finder instead); Sending email.
+
+- ID: `cybermax/email-validate` · MCP tool: `cybermax--email-validate` at https://mcp.apify.com/?tools=cybermax/email-validate
+- HTTP: `POST https://api.apify.com/v2/acts/cybermax~email-validate/run-sync-get-dataset-items` (header `Authorization: Bearer <APIFY_TOKEN>`, body = input JSON, returns a JSON array)
+- Price: $0.0008 per email validated (pay per event, billed by Apify to the caller's account)
+- Store page: https://apify.com/cybermax/email-validate
+- Input fields: `emails` (array, required), `suggestTypos` (boolean), `checkDomainAge` (boolean), `checkWebsite` (boolean), `maxConcurrency` (integer), `timeoutSecs` (integer)
+- Output fields: `input`, `email`, `verdict`, `score`, `reasons`, `summary`, `localPart`, `domain`, `syntax`, `domainExists`, `mx`, `flags`, `suggestion`, `domainHealth`, `catchAll`, `mailbox`, `errors`, `ms`, `checkedAt`
+- One record per input item; invalid or unresolvable inputs are returned with an error field and are not charged unless the pricing says otherwise.
+- On the hosted MCP server the tool is named "cybermax--email-validate" and takes the same input object (plus optional waitSecs).
+
+Example input:
+
+```json
+{"emails":["jane.doe@gmail.com","info@stripe.com","someone@mailinator.com","bob@gmial.com","test@this-domain-does-not-exist-9x7q.com"]}
+```
+
+Python:
+
+```python
+from apify_client import ApifyClient  # pip install apify-client
+client = ApifyClient("<APIFY_TOKEN>")
+run = client.actor("cybermax/email-validate").call(run_input={"emails":["jane.doe@gmail.com","info@stripe.com","someone@mailinator.com","bob@gmial.com","test@this-domain-does-not-exist-9x7q.com"]})
+for item in client.dataset(run["defaultDatasetId"]).iterate_items():
+    print(item)
+```
+
+MCP client config:
+
+```json
+{
+  "mcpServers": {
+    "cybermax-email-validate": {
+      "type": "http",
+      "url": "https://mcp.apify.com/?tools=cybermax/email-validate",
+      "headers": {
+        "Authorization": "Bearer <APIFY_TOKEN>"
+      }
+    }
+  }
+}
+```
+
+Catalog: https://shaw276-cybermax.static.hf.space/market/catalog.json · All tools: https://shaw276-cybermax.static.hf.space/market/llms.txt

market/tools/html-to-pdf.md

@@ -4,7 +4,11 @@
 
 **Status: coming soon, not callable yet.**
 
-HTML to PDF and URL to PDF API with headless Chrome: A4/Letter/Legal, landscape, margins, header/footer with page numbers, print CSS. Returns a public PDF link, page count and size. Schedule it to archive web pages as PDFs. $0.003 per PDF, failures free.
+Printwren. HTML to PDF and URL to PDF with headless Chrome: give web page URLs and/or one complete raw HTML document (invoice, report, receipt, ticket) and it returns, per document: pdfUrl (public link that opens without a token), pages (page count), bytes, page title, finalUrl after redirects and httpStatus, plus the paper size used; failed documents come back in "failed" with the reason. Options: A4/Letter/Legal/A3/A5, landscape, margins, page numbers, HTML header/footer templates, print or screen CSS, wait for a selector, PDF bookmarks from headings, tagged PDF, optional base64 bytes. Cookie banners are dismissed automatically. Use it to render an invoice or report you just generated as a PDF, archive a web page as a dated PDF, or attach a printable copy of an article to an email or ticket. Not for screenshots (use SnapScout screenshot_website), Word/Excel/PowerPoint files (use Paperkiln office_to_pdf), or pages behind a login or CAPTCHA. Cost: $0.003 per PDF generated; failed pages are free.
+
+Use when: render an invoice or report you just generated as a PDF, archive a web page as a dated PDF, or attach a printable copy of an article to an email or ticket.
+
+Not for: Not for screenshots (use SnapScout screenshot_website), Word/Excel/PowerPoint files (use Paperkiln office_to_pdf), or pages behind a login or CAPTCHA.
 
 - ID: `cybermax/html-to-pdf` · MCP tool: `cybermax--html-to-pdf` at https://mcp.apify.com/?tools=cybermax/html-to-pdf
 - HTTP: `POST https://api.apify.com/v2/acts/cybermax~html-to-pdf/run-sync-get-dataset-items` (header `Authorization: Bearer <APIFY_TOKEN>`, body = input JSON, returns a JSON array)

market/tools/office-to-pdf.md

@@ -4,7 +4,11 @@
 
 **Status: coming soon, not callable yet.**
 
-Convert DOCX, XLSX, PPTX, ODT, RTF, CSV and legacy DOC/XLS/PPT files to PDF (or PDF/A) in bulk with headless LibreOffice and Office-compatible fonts. Give URLs or share links (Dropbox, Google Drive, OneDrive); get hosted PDFs with page count. $0.005 per file; failures free.
+Paperkiln. Converts office documents to PDF (or archival PDF/A-2b) with headless LibreOffice and Office-compatible fonts: .docx .doc .xlsx .xls .pptx .ppt .odt .ods .odp .rtf .csv .txt .vsdx by URL, including public Dropbox, Google Drive, OneDrive/SharePoint and Google Docs/Sheets/Slides share links. Returns, per file: pdfUrl (shareable link that works without a token), pages (page count), pdfBytes, inputFormat, fileName, whether it is PDF/A and conversion time; failed files come back in "failed" with the reason. Use it to turn a generated DOCX or XLSX report into a PDF you can send, batch-convert a folder of contracts to PDF/A for archiving, or get a page count before an e-signature step. For web pages or raw HTML use Printwren html_to_pdf instead (it renders in a real browser). Not for PDF to Word, password-protected files or running macros. Cost: $0.005 per file converted; failed downloads, unsupported files and conversion errors are free.
+
+Use when: turn a generated DOCX or XLSX report into a PDF you can send, batch-convert a folder of contracts to PDF/A for archiving, or get a page count before an e-signature step.
+
+Not for: For web pages or raw HTML use Printwren html_to_pdf instead (it renders in a real browser); Not for PDF to Word, password-protected files or running macros.
 
 - ID: `cybermax/office-to-pdf` · MCP tool: `cybermax--office-to-pdf` at https://mcp.apify.com/?tools=cybermax/office-to-pdf
 - HTTP: `POST https://api.apify.com/v2/acts/cybermax~office-to-pdf/run-sync-get-dataset-items` (header `Authorization: Bearer <APIFY_TOKEN>`, body = input JSON, returns a JSON array)

market/tools/rss-feed-finder.md

@@ -4,7 +4,11 @@
 
 **Status: coming soon, not callable yet.**
 
-Find every RSS, Atom and JSON feed a website publishes (link tags, page links, common paths, platform rules for Reddit, Medium, YouTube, GitHub), validated, with title, type, last post date and the latest items. Monitor mode returns only new posts. $0.002 per feed found.
+Feedpeck. RSS feed finder and reader: give websites (domains, blog sections like stripe.com/blog, subreddits, Medium profiles, YouTube channels, GitHub repos or feed URLs) and it discovers every RSS, Atom and JSON feed they publish (link tags, page links, common paths, platform rules), validates each, and returns per feed: feedUrl, feedTitle, feedType, discoveredVia, lastItemAt (date of the newest post), itemCount and latestItems (title, url, published, summary, image). With output="items" it returns one row per post instead; onlyNew=true returns only posts not seen on an earlier call. Sites with no feed are listed in sitesWithoutFeed. Use it to get the feed URL for a blog, check whether a company or competitor is still publishing (lastItemAt), or pull the latest posts from 20 sites for a digest. Feeds often carry only a summary: for full article text pass the post URLs to Inkpluck extract_article. Cost: $0.002 per feed found (latest posts included), or $0.0003 per post with output="items" or onlyNew; sites without a feed are free.
+
+Use when: get the feed URL for a blog, check whether a company or competitor is still publishing (lastItemAt), or pull the latest posts from 20 sites for a digest.
+
+Not for: With output="items" it returns one row per post instead; onlyNew=true returns only posts not seen on an earlier call.
 
 - ID: `cybermax/rss-feed-finder` · MCP tool: `cybermax--rss-feed-finder` at https://mcp.apify.com/?tools=cybermax/rss-feed-finder
 - HTTP: `POST https://api.apify.com/v2/acts/cybermax~rss-feed-finder/run-sync-get-dataset-items` (header `Authorization: Bearer <APIFY_TOKEN>`, body = input JSON, returns a JSON array)