"How to build your own knowledge base" matching MCP connectors:

For queries a model can't confidently place: resolves or declines. Built to decline, not guess.

A semantic search server that gives AI assistants instant access to Dodo Payments documentation and knowledge base.

**ColdState Knowledge Search MCP Server** https://github.com/daniel-coldstate/coldstate-mcp Semantic search over 64.6M knowledge entries — the structured alternative to web search APIs and web scraping for LLM agents. No crawling, no rate limits, sub-3s responses. Cloud-hosted at services.coldstate.ai

Create your digital twin: a public page that answers from your own material, with citations.

Diffbot MCP — Knowledge Graph company enrichment + web content extraction (diffbot.com)

LLM-ready web search + instant answers + URL-to-clean-text fetch for agents and RAG.

Document-to-Markdown MCP server — convert PDF, Office and HTML into LLM-ready Markdown.

NVIDIA AI knowledge graphs — 20 domains. 4x F1, 11x fewer tokens, SHA-256 provenance. MCP-native.

Build and run grounded business agents over MCP: agents, knowledge bases, skills, Storylines.

Manage your Mistral platform — models, files, batch jobs, agents and RAG document libraries.

Extract PDFs to Markdown, RAG chunks and cited tables; publish tracked Doc Links with read stats.

Parse PDF/Word/PPT/HTML to Markdown; tables as JSON, image extraction, RAG chunking, page ranges.

Your company's brain for AI agents. Cited, permission-aware knowledge across every system.

A multi-domain reference corpus where every answer drills down to its cited, verified source.

Multilingual YouTube → Knowledge Pack engine. Paste a video URL and get a structured pack — summary, key ideas, glossary, quiz, transcript with timestamps — in Spanish, Portuguese, German, or English. Anonymous endpoint plus OAuth-gated tools for library search, RAG Q&A on a single pack, and Anki export.

Semantic search over a 200-chunk ACLM lifestyle medicine knowledge base.

The world's first named AI prompt quality score. Score, optimize, and compare LLM prompts before they hit any model. Free tier available. Built on PEEM, RAGAS, G-Eval, and MT-Bench frameworks. x402-native on Base.

Marketplace where AI agents buy and sell books — curated domain knowledge that improves tasks.

Syracuse is an MCP server that gives agents reliable company and industry/region news. Every result is a structured event that is typed, dated, and linked to its source article. It's built for precision over volume, so an agent can act on it directly without a human in the loop weeding out wrong-entity matches or hallucinated stories. It's free for individuals, and in an open, anonymised benchmark against Exa, Tavily, Linkup and Perplexity it currently leads on company news.

LLM caching proxy (x402 USDC on Base) - exact + semantic cache. Free health.