OpenLore MCP Server已验证已精选

What is OpenLore?

Serve your docs to AI agents over SSH and MCP. Agents use ls, cat, grep, find, pipes, and shell loops against a virtual filesystem backed by your documentation. No ingestion pipeline, no vector DB. Point it at a directory and it serves the existing hierarchy directly.

How it works

Agent ──SSH or MCP──▶ OpenLore ──▶ docs, knowledge, and artifacts

By default it starts SSH on localhost:2222, a web view on http://localhost:8080, and MCP over HTTP on http://localhost:8080/mcp.

Key features

• Agent-native retrieval — agents use the shell tools they already know instead of learning a bespoke API
• One knowledge surface, multiple transports — SSH, SFTP/SSHFS, MCP, and a human-friendly web view
• Identity-scoped views — each agent sees only the docsets it needs, with role-based access
• Governed writing — writes are atomic, conflict-aware, and can require human approval
• Safe by construction — the shell is an in-memory Go interpreter, not a real OS shell. No escape, no arbitrary execution
• Portable knowledge bundles — embed docs into a self-contained binary or package as a desktop MCP extension

Use cases

• Internal API docs, runbooks, and architecture notes for coding agents
• Shared live memory for teams of agents across sessions
• Agent Skills publishing into governed, discoverable collections
• Remote review of agent-generated reports and artifacts through SSH or browser

Quick start

go install github.com/aakarim/go-openlore/cmd/openlore@latest
openlore ./docs
ssh -p 2222 localhost