MidOS - MCP Community Library MCP Server
io.github.MidOSresearch/midos
Curated knowledge base for AI agents — 19K chunks, 121 skills, semantic search via MCP
What is the MidOS - MCP Community Library MCP server?
The MidOS MCP Community Library is an MCP server that exposes a curated knowledge base of 19K+ chunks, 121 skills, and research patterns to AI agents like Claude and Cursor. It provides semantic search, verified improvements (EUREKA), state-of-the-art benchmarks, and reusable patterns across 16+ technology stacks, with optional vector search via LanceDB.
MidOS gives your AI agent access to a structured, searchable knowledge base covering domains from security ops to software architecture. It includes 119 reusable skills, 383 validated improvements with measured ROI, and 140 state-of-the-art convergence ceilings. Free tier offers 500 queries/month with full content; Pro tier ($10/mo) adds 100K queries and advanced tools.
How to install MidOS - MCP Community Library
Copy-paste configuration for popular MCP clients.
MIDOS_API_KEYsecretMidOS API key for authenticated access (optional for community tier)
Tools & capabilities
Tools this server exposes to the agent.
Knowledge Search— Search 46K+ curated chunks across all domains with full content (no truncation)Skills Lookup— Access 119 reusable patterns across 16+ technology stacksEUREKA Improvements— Browse 383 validated improvements with measured ROITruth Patches— Query 50 verified patches and correctionsSOTA Benchmarks— Explore 140 state-of-the-art convergence ceilingsVector Search— Semantic search via LanceDB + Gemini embeddings (3072-d)
Use cases
- Search curated knowledge base to answer architecture and design questions for your AI agent
- Access reusable skills and patterns to accelerate development across multiple tech stacks
- Retrieve validated improvements (EUREKA) with measured ROI to optimize systems
- Query state-of-the-art benchmarks and convergence ceilings for performance tuning
- Perform semantic search on 46K+ chunks to find contextually relevant solutions
- Integrate verified patches and corrections into your AI agent's decision-making
MidOS - MCP Community Library MCP server FAQ
MidOS is an MCP server that exposes a curated knowledge base of 19K+ chunks, 121 skills, research patterns, and verified improvements to AI agents. It supports both keyword and semantic search via LanceDB embeddings.
Yes. The Dev tier is free and includes 500 queries/month with full content access. Pro tier costs $10/mo and offers 100K queries/month plus advanced tools (security ops, AOTC, orchestration).
Install via pip (pip install midos-mcp), set MIDOS_ROOT to your MidOS directory, then run `midos-mcp config --generate cursor` or `--generate claude` to get ready-to-paste JSON config for your client.
No API key is required for the free Dev tier. Pro tier requires a paid API key generated via `midos-mcp keys generate`. Invalid or expired keys silently fall back to Dev tier.
MidOS supports stdio (default for Claude Code, Cursor, Cline), HTTP (for web clients and remote access), and SSE (legacy clients). Configure via `midos-mcp serve --stdio`, `--http`, or `--sse`.
Python 3.10+, FastMCP 2.x, and a local MidOS knowledge base. The knowledge base is auto-detected or configured via the MIDOS_ROOT environment variable.
README (reference)
Source of truth, from the repository.
midos-mcp
MidOS MCP Community Library — Knowledge, Skills, and Research as MCP tools.
Install once, connect to any AI client.
Install
pip install midos-mcp
Requirement: MidOS knowledge base must be available locally. Set the MIDOS_ROOT environment variable to point to your MidOS installation directory (the one containing knowledge/ and modules/).
export MIDOS_ROOT=/path/to/midos
Quick Start
# Start server (stdio — default for Claude Code, Cursor, etc.)
midos-mcp serve
# Start with HTTP transport
midos-mcp serve --http
# Generate config for your AI client
midos-mcp config --generate claude
midos-mcp config --generate cursor
# Check health
midos-mcp health
Commands
midos-mcp serve
Start the MCP server with your preferred transport.
midos-mcp serve # stdio (default)
midos-mcp serve --http # Streamable HTTP on 127.0.0.1:8419
midos-mcp serve --sse # Legacy SSE transport
midos-mcp serve --http --host 0.0.0.0 --port 9000
| Flag | Transport | Use Case |
|---|---|---|
--stdio | stdio | Claude Code, Cursor, Cline (default) |
--http | Streamable HTTP | Web clients, remote access |
--sse | Server-Sent Events | Legacy clients |
midos-mcp config
Generate ready-to-paste JSON config for any supported MCP client.
midos-mcp config --generate claude # Claude Desktop / Claude Code
midos-mcp config --generate cursor # Cursor IDE
midos-mcp config --generate cline # Cline (VS Code)
midos-mcp config --generate windsurf # Windsurf IDE
midos-mcp config --generate continue # Continue.dev
midos-mcp config --generate zed # Zed Editor
midos-mcp config --generate opencode # OpenCode
midos-mcp config --generate http # Generic HTTP client
# Write to file
midos-mcp config --generate claude -o mcp-config.json
midos-mcp health
Check server health: knowledge base stats, vector store status, dependencies.
midos-mcp health # Human-readable output
midos-mcp health --json # JSON output for scripts
midos-mcp keys
Manage API keys for tiered access control.
midos-mcp keys generate --name "my-app" --tier dev
midos-mcp keys list
midos-mcp keys revoke midos_sk_abc
Tiers: dev (free), pro ($10/mo), admin.
Configuration
Environment Variables
| Variable | Default | Description |
|---|---|---|
MIDOS_ROOT | auto-detect | Path to MidOS installation (required) |
MIDOS_HOST | 127.0.0.1 | HTTP server bind host |
MIDOS_PORT | 8419 | HTTP server bind port |
MIDOS_TRANSPORT | stdio | Default transport |
Root Detection
The server locates the MidOS knowledge base:
MIDOS_ROOTenvironment variable (recommended)- Auto-detect: walks up from package location looking for
CLAUDE.md+knowledge/ - Fallback: current working directory
If you installed via pip, auto-detect won't find the knowledge base. Either:
# Option A: Clone the repo and point to it
git clone https://github.com/MidOSresearch/midos.git
export MIDOS_ROOT=./midos
# Option B: Point to an existing MidOS directory
export MIDOS_ROOT=/path/to/your/midos
Verify with midos-mcp health — it shows the detected root and knowledge counts.
Supported Clients
| Client | Config File | Transport |
|---|---|---|
| Claude Desktop | claude_desktop_config.json | stdio |
| Claude Code | .mcp.json | stdio |
| Cursor | .cursor/mcp.json | stdio |
| Cline | cline_mcp_settings.json | stdio |
| Windsurf | ~/.codeium/windsurf/mcp_config.json | stdio |
| Continue.dev | ~/.continue/config.json | stdio |
| Zed | ~/.config/zed/settings.json | stdio |
| OpenCode | opencode.json | SSE |
Optional Dependencies
# Vector store support (LanceDB + embeddings)
pip install midos-mcp[vector]
# Development tools
pip install midos-mcp[dev]
What's Inside
MidOS exposes 68 MCP tools across 2 tiers:
| Tier | Tools | Access |
|---|---|---|
| Dev | 34 tools | Free — no API key needed, full content, 500 queries/month |
| Pro | 68 tools | $10/mo — security ops, AOTC, orchestration, maker write ops |
Capabilities
- Knowledge — Search 46K+ curated chunks across all domains (full content, no truncation)
- Skills — 119 reusable patterns across 16+ technology stacks
- EUREKA — 383 validated improvements with measured ROI
- Truth — 50 verified patches and corrections
- SOTA — 140 state-of-the-art convergence ceilings
- Vector Search — Semantic search via LanceDB + Gemini embeddings (3072-d)
Rate Limits
| Tier | Queries/month | Content |
|---|---|---|
| Dev | 500 | Full content (no truncation) |
| Pro | 100,000 | Full content + AOTC + ops packs |
Invalid or expired API keys silently fall back to Dev tier.
Requirements
- Python 3.10+
- FastMCP 2.x
- MidOS knowledge base (local)
License
MIT
Related MCP servers

MidOS - MCP Community Library
Curated knowledge API for AI agents—skill packs, semantic search, and validated patterns.

Memory Vault
Self-hosted AI memory layer with hybrid search, knowledge graphs, and MCP integration for Claude.
Generate design token systems from a hex color — OKLCH-derived, WCAG-validated, multi-format
View repository →
ANIME INTELLIGENCE
Anime shopping intelligence: identity, value, authenticity, BUY/WAIT and purchase routing via x402.

Meta Ads MCP by ScaleForge — control Facebook & Instagram Ads via Graph API v24.0. 32 tools.
Live tools for USDi: CPI-indexed exchange rate, contract/pool addresses, and mint/redeem mechanics.

