"OpenAI's Web Retrieval Infrastructure" matching MCP connectors:
GET /v1/connectors – MCP directory API referenceMatching Connector Tools:
Engram is a persistent, long-term memory layer for AI agents and assistants. Claude, ChatGPT, Grok, Cursor and any MCP client share one memory, stored as plain markdown notes: your knowledge base, second brain and AI context in one place. No extraction step: the memory is the note itself, so you can read exactly what your AI remembers and fix it. Edit your memory in Obsidian (real-time sync), the web app, or on your phone. Hybrid keyword + semantic search (RAG over your notes) finds exact strings like error messages, config keys and IDs. Remote MCP server over Streamable HTTP with OAuth 2.1; notes encrypted at rest. Https://engram.page https://youtu.be/rwnPeZ-8Lqo?is=NI-N7BduydGAiZlF https://github.com/engram-app/Engram
**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
Direct access to 60+ scraping and search tools. Extract structured data from Google (Search, Maps, Trends), Amazon, Airbnb, Social Media, and any web page directly into your AI agent.
Search Fragments — two tools for the queries an agent can't place, both built to decline rather than guess. resolve_fragment takes a half-remembered, cross-source query ("a musician who became famous for stopping performing") and returns a grounded answer, ranked web sources to confirm by eye, or an explicit no-resolution. verify_claim takes a specific factual assertion and returns supported, partially_supported, insufficient_evidence, or unsupported, with cited evidence and a stated_limits field that is always present. There is no confidence score — insufficient_evidence fires freely, and unsupported requires a source that explicitly contradicts, never mere absence of confirmation. Every verdict is decide-by-eye: "supported" means current web sources confirm it, not that the claim is true. Calibrated against 18 known claims before release. Free, no signup. Streamable HTTP (MCP 2025-11-25). Read-only.
Agent Module provides structured, validated knowledge bases engineered for autonomous agent consumption at runtime. Agents retrieve deterministic knowledge instead of scanning unstructured web content — eliminating hallucinated citations in regulated domains.
Web search for AI agents. Ranked results with page passages already extracted, plus URL to markdown.
Fact-check claims against live web sources: verdict, consensus, cited sources, PII redaction.
Semantic search across 50,000+ food recipes with hybrid retrieval and reranking.
Market data and web intelligence for AI agents, paid per call in USDC on Base via x402.
Knowledge Base von designare.at – Michael Kanda, Web & KI aus Wien. Semantische Suche über RAG.
Semantic search over a 200-chunk ACLM lifestyle medicine knowledge base.
AI web extraction: send URLs + a JSON Schema, get clean structured data. Pay-per-use via x402.
MCP-native web evidence and claim verification: cited, source-grounded evidence for AI agents.
Web-scale search for AI thats 100x cheaper and 10x faster. https://www.ceramic.ai/
Exa MCP — neural/semantic web search + content retrieval (exa.ai)
Diffbot MCP — Knowledge Graph company enrichment + web content extraction (diffbot.com)
Brave Search MCP — independent web index (no Google/Bing dependency)
Search 18M+ legal documents across 110+ countries. Case law, legislation, and doctrine with semantic + keyword hybrid search. Supports tool discovery, multi-jurisdictional queries, citation resolution, and full document retrieval. Requested missing datasets can be fully indexed within 48h.
Memwyre is an MCP-native persistent memory layer for AI agents, synchronizing context across Claude Code, Cursor, VS Code, and OpenClaw. Built with a high-precision retrieval architecture (dense vector search, BM25, and cross-encoder reranking), Memwyre achieves a benchmarked 73.1% accuracy on the Long-Context Memory (LoCoMo) benchmark. It provides secure, isolated knowledge vaults with dedicated tools (search_memwyre, save_memory, list_memories) to save and recall structured project decisions.
Universal memory runtime for AI agents: episodic, semantic, and procedural memory with hybrid retrieval and spaced-repetition decay.
