ld-spec-mcp
Verified Safeby robsyc
Overview
Serves W3C Semantic Web specifications section-by-section and resource-by-resource to AI agents for efficient, targeted information retrieval.
Installation
python src/main.pyEnvironment Variables
- SPEC_VERSIONS
- CACHE_TTL
Security Notes
The server fetches content from a predefined set of trusted W3C URIs listed in `index.yaml`. User input (`spec_key`, `ns_key`) is used to look up these trusted URIs, preventing arbitrary URL fetching (SSRF). While fetching external content inherently carries some risk, robust libraries like `httpx` (with timeout) and `BeautifulSoup`/`RDFLib` are used, along with explicit sanitization in `html_to_markdown` for known `html-to-markdown` library issues. No direct command injection, use of `eval`, or hardcoded secrets were identified. Input validation relies on `FastMCP` framework's `Annotated` types.
Similar Servers
mcp-framework-server
A Python-based server for a Model Context Protocol (MCP) enabling interactive career orientation (proforientation) dialogues via a REST API, designed to integrate with a Telegram bot.
mcp-fastmcp-starter
Provides a minimal, production-minded MCP-shaped tool server for building agentic AI services with typed tool contracts and deterministic JSON I/O via FastAPI.
MCP_Server_Project
A foundational FastAPI server for integrating and managing Spark and PostgreSQL data operations and ETL pipelines.
ai-mcp-server-base
Provides a base server for building AI Microservice Orchestration applications using FastMCP and FastAPI, featuring tools and resources for AI agents.