Firecrawl MCP ServerMCP ServersA Firecrawl connector for web search, page extraction, site mapping and bounded crawling with Markdown or structured results.Open sourceSelf-hostedFreeViews · 7 daysTop 2%
Research, Knowledge & Docs MCP Servers
Research, Knowledge & Docs mcp servers for AI agents and assistants.
7 results in Research, Knowledge & Docs MCP Servers
Active Pro listings from this category appear first. Your sort applies within Pro and Free results.
Firecrawl MCP ServerMCP ServersA Firecrawl connector for web search, page extraction, site mapping and bounded crawling with Markdown or structured results.Open sourceSelf-hostedFree19 views7,769 stars
itsnotesNotes & Knowledge ManagementMIT-licensed, self-hosted note board in the style of Google Keep, run with Docker and PostgreSQL. Adds month markers, folders, saved searches, note history, Keep and Obsidian import, a Markdown mirror, optional AI and a read-only MCP server.Open sourceSelf-hostedViews · 7 daysNo views in 7 days
MCP Brave Search ServerMCP ServersAn archived Brave Search API connector for web results and local business searches, with pagination and filtering.Open sourceSoftwareFreemiumViews · 7 daysNo views in 7 days
MCP Fetch ServerMCP ServersA lightweight Python MCP connector that retrieves a URL and converts HTML into Markdown, with chunked output for long pages.Open sourceSoftwareFreeViews · 7 daysNo views in 7 days
MCP Memory ServerMCP ServersA local knowledge-graph memory server that stores named entities, directed relations and observations for later retrieval.Open sourceSoftwareFreeViews · 7 daysNo views in 7 days
Views · 7 daysNo views in 7 days
ZenNotes & Knowledge ManagementZen is an AGPL-licensed, single-user notes server: Markdown stored in SQLite, tags and Focus Modes instead of folders, BM25 search, version history, templates, tag-scoped API tokens and an experimental MCP server. It runs from Docker or one Go binary.Open sourceSelf-hostedViews · 7 daysTop 29%
Shortlists for Research, Knowledge & Docs MCP Servers
Explore collections with tools relevant to this category, grouped by deployment, source model or workload.
Best MCP servers by workflow →
2 matching tools from this category
Free IT tools →
4 matching tools from this category
Collaboration tools with APIs for developer teams →
2 matching tools from this category
How to choose research, knowledge & docs mcp servers tools
Research, knowledge and docs MCP servers let AI clients search the web, fetch and clean pages, crawl sites, and read or write personal knowledge stores such as note vaults and memory graphs. They give an assistant fresh information beyond its training data and a place to keep facts between sessions. They differ from business servers because the sources are public content or a personal knowledge base rather than shared workplace systems.
Who uses these tools
- Researchers gathering sources inside an assistant
- Writers grounding drafts in current web pages
- Note-takers querying a personal knowledge vault
Capabilities to compare
- Web search API coverage and result quality
- Page fetching with HTML to Markdown conversion
- Site crawling depth and rate controls
- Note vault read, search and edit tools
- Persistent memory of entities and relations
Operating considerations
- Search API keys and per-query cost
- Prompt injection hidden in fetched pages
- Robots rules and site terms when crawling
- Edits and deletions in your note vault
- Where memory data is stored locally
Selection checklist
- Decide whether you need search, single-page fetching, full crawling or note access, and pick servers that match only that scope.
- Check how fetched content is cleaned and chunked, since long raw pages waste context and reduce answer quality.
- Disable delete and overwrite tools on note servers until you trust how the assistant edits files.
- Treat fetched web text as untrusted input and confirm the client asks before acting on instructions found in pages.
What belongs in this category: A server is included when it retrieves external content, converts pages for model consumption, or stores and recalls knowledge for an AI client, rather than operating code, infrastructure or business applications.
Research, Knowledge & Docs MCP Servers on iThub in figures
iThub lists 7 tools in Research, Knowledge & Docs MCP Servers. 7 of 7 are open source, 3 of 7 can be self-hosted, 7 of 7 have a free, freemium or open-source pricing model and 2 of 7 offer an API.
- Tools listed
- 7
- Open source
- 7
- Self-hostable
- 3
- Free to start
- 7
- With an API
- 2
- Web app
- 2
- Desktop app
- 2
Recently added: itsnotes, Zen, MCP Obsidian Server, Firecrawl MCP Server
Typical use cases
- Collecting and summarizing sources for a technical comparison
- Converting documentation pages into Markdown for an agent
- Searching a personal note vault from inside a chat client
- Remembering user preferences and project facts across conversations
Trade-offs to model
- Hosted crawling services handle rendering and blocking for you, while local fetch servers are free but fail on script-heavy pages.
- A memory server makes assistants feel consistent, but stored facts can go stale and need regular review or pruning.
Common mistakes
- Letting an agent follow instructions embedded in fetched pages, which turns untrusted web content into commands.
- Running a crawler without rate limits and getting your address blocked by the sites you rely on.
Browse all MCP servers →Browse development MCP servers →See self-hosted knowledge base software →
Decision resources
Continue the technical evaluation
Use the resources that match your decision stage. Profile data and linked sources were refreshed through November 7, 2026.
0 tools selectedCompare selected