{
  "markdown": "# Fact Verification MCP\n\n**Fact verification for AI agents** — give it a claim, get a **verdict**\n(`supported` / `disputed` / `unverifiable`) with a **0-100 confidence** and **cited\nsources**. It classifies each claim's domain and cross-references the FoundryNet\nData Network plus a general web search.\n\n> Part of the **FoundryNet Data Network**. Verification results carry a verifiable\n> provenance attestation. See also:\n> **financial-signals-mcp**, **cyber-intel-mcp**, **patent-intel-mcp**,\n> **gov-contracts-mcp**, **compliance-mcp**, **brand-intel-mcp**,\n> **weather-intel-mcp**, **academic-intel-mcp**, **oss-intel-mcp**,\n> **social-intel-mcp**.\n\n## Connect\n\n- **MCP endpoint** (Streamable HTTP): `https://fact-check-mcp-production.up.railway.app/mcp`\n- **Registry:** `io.github.FoundryNet/fact-check-mcp`\n- **Agent card:** `https://fact-check-mcp-production.up.railway.app/.well-known/agent-card.json`\n\n### Claude Desktop / Cursor / Claude Code\n\n```bash\nclaude mcp add --transport http fact-check https://fact-check-mcp-production.up.railway.app/mcp\n```\n\n```json\n{ \"mcpServers\": { \"fact-check\": { \"url\": \"https://fact-check-mcp-production.up.railway.app/mcp\" } } }\n```\n\n## Tools\n\n| Tool | Price | What it does |\n|---|---|---|\n| `verify_claim` | $0.02 | Verify a claim → verdict, confidence, cited sources, explanation |\n| `batch_verify` | $0.01/claim (min $0.05) | Verify many claims in one call |\n| `source_check` | $0.01 | Assess a source URL — domain age (RDAP), trust signals, bias indicators |\n| `daily_brief` | $5 | Curated daily brief — top disputed/most-verified claims, trending topics |\n| `brief_summary` | $0.50 | Structured top-5 signals from today's brief — a cheap sample of `daily_brief` |\n| `mint_info` | **free** | FoundryNet Data Network attestation info |\n\n**Free tier:** 10 verifications/day per agent. Then metered: the tool returns an\nHTTP-402 payment challenge — pay it, re-call with the same args plus\n`payment_tx=<signature>`. An `Authorization: Bearer fnet_…` key bypasses the paywall.\n\n## How it works\n\nEach claim is classified by domain and cross-referenced against the most relevant\nFoundryNet data source **and** a general web search:\n\n- **company** claims → `brand-intel-mcp`\n- **finance** claims → `financial-signals-mcp`\n- **patents** claims → `patent-intel-mcp`\n- **regulation** claims → `compliance-mcp`\n\nThe verdict is synthesized from source agreement and returned with the citations\nthat drove it. Results are **cached for 24h** (keyed by a normalized-claim hash) and\ncarry a **verifiable provenance attestation** so a buyer can verify the result was\nproduced by this server, unaltered.\n\n> The general web-search backend needs a `WEB_SEARCH_API_KEY` (Brave Search API).\n> Without it, claims with no FoundryNet corroboration return `unverifiable` with a\n> note rather than a guess.\n\n## Sources\n\nOn demand: a general **web search** (Brave Search API) + the **FoundryNet Data\nNetwork** sibling servers (called over their public MCP endpoints). Source-credibility\nchecks use keyless **RDAP** for domain age. Cached in a standalone Supabase project.\n\n## Discovery\n\nMCP registry: `io.github.FoundryNet/fact-check-mcp`\n\nBuilt by [FoundryNet](https://foundrynet.io?utm_source=github&utm_medium=readme&utm_campaign=fact-check-mcp) · forge@foundrynet.io\n\n## Live network activity\n\nReal-time verified work across 17 servers and autonomous agents. See [foundrynet.io](https://foundrynet.io?utm_source=github&utm_medium=readme&utm_campaign=fact-check-mcp).\n",
  "bytes": 3490,
  "sha": "35eb945d56a69dc39d1d531b0a0aeddc9b1b074e6108b4335405b9b6a57e1f5c",
  "repo_slug": "foundrynet/fact-check-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_foundrynet_fact_check_mcp_fe0f8fbd/readme"
}