{
  "markdown": "# MCP Fact-Check MCP Server\n\nAn MCP Server for validating code or content against the official **Model Context Protocol (MCP)** specification to ensure technical accuracy and prevent the spread of misinformation.\n\n📦 **[View in MCP Registry](https://registry.modelcontextprotocol.io/?q=mcp-factcheck)** - Available in the official MCP Registry\n\n📋 **[View Project Roadmap](ROADMAP.md)** - See planned features and development progress\n\n🏗️ **[Design Documentation](DESIGN.md)** - Technical design and implementation details\n\n## Overview\n\nThe MCP Fact-Check MCP Server helps ensure technical accuracy when coding or writing about MCP by comparing content against official specifications. It uses:\n\n- **Semantic search** with OpenAI embeddings to find relevant specification sections\n- **AI-powered validation** to detect inaccuracies and suggest corrections\n- **Compound claim decomposition** to validate complex statements with multiple assertions\n- **Multiple spec versions** support (draft, 2025-06-18, 2025-03-26, 2024-11-05)\n\n## Features\n\n### MCP Tools Exposed\n\n1. **`check_mcp_claim`** - Comprehensive validation of MCP-related content\n\n   - Validates multi-claim content (documentation, tutorials, bullet points)\n   - Automatically decomposes compound claims (e.g., \"X and Y\") for accurate validation\n   - Provides step-by-step validation workflow\n   - Identifies missing best practices and modal verb issues\n   - Returns corrected content with confidence scores\n\n2. **`check_mcp_quick_fact`** - Quick fact-checking for single MCP claims\n\n   - Validates single sentences or quick questions\n   - Returns concise ✓/✗ verdict with explanation\n   - Uses aggressive search strategies for accuracy\n   - Perfect for \"Does MCP support X?\" questions\n\n3. **`search_spec`** - Searches MCP specifications using semantic similarity\n\n   - Returns most relevant specification sections\n   - Supports all specification versions\n\n4. **`list_spec_versions`** - Lists available MCP specification versions\n   - Shows version dates and descriptions\n   - Indicates which version is current\n\n### MCP Prompts Available\n\n1. **`migrate-mcp-content`** - Guides content migration between MCP specification versions\n\n   - Validates content against source specification first\n   - Identifies changes between specification versions\n   - Provides step-by-step migration guidance\n   - Works with any type of MCP-related content\n   - Preserves the original tone, style, and voice when making corrections or suggestions\n\n   **Parameters:**\n\n   - `current_version` (required): Source MCP specification version (e.g., \"2024-11-05\", \"2025-06-18\")\n   - `target_version` (required): Target MCP specification version to migrate to (e.g., \"draft\")\n   - `update_scope` (optional): Determines how aggressive the migration should be\n     - `critical_only`: Fix only critical inaccuracies and breaking changes (minimal changes)\n     - `enhancement_focused`: Fix issues and improve clarity, align with best practices\n     - `comprehensive`: Complete review with all improvements and enhanced clarity\n     - Default: `comprehensive`\n\n## Installation\n\nThe MCP Fact-Check server is available through the [Model Context Protocol registry](https://github.com/modelcontextprotocol/servers). Install it directly from your MCP client:\n\n**For Claude Desktop and other MCP clients:**\n\n- Search for \"mcp-factcheck\" in your client's server marketplace\n- Click install\n- Provide your OpenAI API key when prompted\n\nThat's it! The server will be automatically configured and ready to use.\n\n> **For developers:** If you need to build from source or contribute to the project, see [INSTALL.md](INSTALL.md) for development setup instructions.\n\n### Observability\n\n#### Visual Tracing with Arize Phoenix\n\nFor a beautiful, AI-focused trace visualization UI, set up Arize Phoenix:\n\n1. **Install and start Phoenix:**\n\n```bash\n# Install Phoenix\npipx install arize-phoenix\n\n# Start Phoenix server\nphoenix serve\n```\n\n2. **Update the Host config to send traces to Phoenix:**\n\n```json\n{\n  \"mcpServers\": {\n    \"mcp-factcheck\": {\n      \"command\": \"/path/to/bin/mcp-factcheck-server\",\n      \"args\": [\n        \"--data-dir\",\n        \"/path/to/data/embeddings\",\n        \"--telemetry\",\n        \"--otlp-endpoint\",\n        \"http://localhost:6006\"\n      ],\n      \"env\": {\n        \"OPENAI_API_KEY\": \"your-api-key\"\n      }\n    }\n  }\n}\n```\n\n3. **View traces at:** http://localhost:6006\n\n**What you'll see in Phoenix:**\n\n- Beautiful AI-focused interface designed for LLM applications\n- Complete validation pipeline timeline with clear visual hierarchy\n- Embedding generation performance and OpenAI API call tracking\n- Vector search visualization with similarity scores\n- Per-chunk validation confidence levels and quality metrics\n- Cost tracking for OpenAI API usage (or whichever llm is being used for embedding the input content/code)\n- Clean, intuitive navigation focused on AI workflows\n\nPhoenix is specifically designed for AI/ML observability and provides a much more user-friendly experience than traditional tracing tools.\n\n## Development\n\n### Building\n\n```bash\n# Build all components\ngo build -o bin/mfc ./cmd/server\ngo build -o bin/specloader ./utils/cmd\n\n# Run tests\ngo test ./...\n```\n\n### Updating Specifications\n\nThe project includes pre-extracted MCP specifications and embeddings for all versions. To check when the draft specification was last updated, see `data/SPEC_METADATA.json`:\n\n```bash\n# View draft update information\ncat data/SPEC_METADATA.json | jq '.specs.draft'\n```\n\n**To update the draft specification:**\n\n```bash\n./bin/specloader spec --version draft\n./bin/specloader embed --version draft\n./bin/specloader embed --version draft-fine\n```\n\n**To add a new specification version:**\n\n```bash\n./bin/specloader spec --version 2025-12-15\n./bin/specloader embed --version 2025-12-15\n./bin/specloader embed --version 2025-12-15-fine\n```\n\nAll specification extraction dates and source commits are automatically tracked in `data/SPEC_METADATA.json`.\n\n### Testing Tools\n\nTest the server using the included test client:\n\n```bash\n# Build test client\ngo build -o bin/factcheck-curl ./cmd/factcheck-curl\n\n# Test tools\n./bin/factcheck-curl --cmd ./bin/mfc --data-dir ./data/embeddings tools/list\n./bin/factcheck-curl --cmd ./bin/mfc --data-dir ./data/embeddings tools/call validate_content '{\"content\":\"MCP is a protocol\"}'\n\n# Test prompts\n./bin/factcheck-curl --cmd ./bin/mfc --data-dir ./data/embeddings prompts/list\n\n# Get migration prompt with minimal parameters\n./bin/factcheck-curl --cmd ./bin/mfc --data-dir ./data/embeddings prompts/get migrate-mcp-content '{\"current_version\":\"2024-11-05\",\"target_version\":\"draft\"}'\n\n# Get migration prompt with all parameters\n./bin/factcheck-curl --cmd ./bin/mfc --data-dir ./data/embeddings prompts/get migrate-mcp-content '{\n  \"current_version\": \"2024-11-05\",\n  \"target_version\": \"2025-06-18\",\n  \"update_scope\": \"critical_only\"\n}'\n```\n\n## Architecture\n\nSee [DESIGN.md](DESIGN.md#architecture-overview) for the complete architecture documentation.\n\n## Environment Variables\n\n- `OPENAI_API_KEY` - Required for embedding generation and content validation\n- `GITHUB_TOKEN` - Optional, for higher GitHub API rate limits when extracting specs\n\n## License\n\nMIT License. See [LICENSE](LICENSE) for details.\n",
  "bytes": 7272,
  "sha": "5e4dd51e00ac1a7306be674ab35fe444f9202d0147a27c9172e2192d5c60c198",
  "repo_slug": "carlisia/mcp-factcheck",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_carlisia_mcp_factcheck_ce8b9a16/readme"
}