litecontext-cli
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@litecontext-clisearch my Obsidian notes for anything about MCP servers"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
LiteContext CLI
LiteContext CLI is a local-first context engine and MCP server for personal knowledge. It indexes local Markdown or Obsidian folders, preserves source citations, and lets compatible AI clients retrieve relevant local context through MCP.
It runs without a desktop application, browser, cloud account, or localhost UI.
Your sources and indexes remain in ~/.litecontext by default.
Install
The Go engine is currently a read-only compatibility preview. It opens the existing LiteContext v1 JSON workspace, applies Context Profile boundaries, and provides cited search, doctor, and MCP stdio without Node at runtime:
go run ./cmd/litecontext doctor --workspace ~/.litecontext --json
go run ./cmd/litecontext search --workspace ~/.litecontext --query "project roadmap" --json
go run ./cmd/litecontext mcp serve --workspace ~/.litecontextIt does not yet write sources, profiles, or SQLite. The TypeScript CLI remains the compatibility baseline for those operations during migration. The first Go binary release will be distributed through GitHub Releases and Homebrew; npm is not a required product runtime.
Related MCP server: ragi
Current TypeScript CLI baseline
litecontext source add ~/Documents/notes --source-id notes --name "Personal notes"
litecontext source list
litecontext doctor
litecontext serve --mcpFor an Obsidian vault, add --source-type obsidian. source add creates a
default Context Profile unless --no-default-profile is supplied. Use
source sync <id> to refresh a source and source remove <id> to remove only
LiteContext's local index and registry entries, never the original files.
MCP configuration
Print a pasteable configuration for an AI client:
litecontext mcp-config --globalThe server command is litecontext serve --mcp. The MCP server reads the same
local workspace when no desktop application is running.
Workspace and architecture
Local Markdown / Obsidian
-> connectors and ContextItem normalization
-> SQLite/FTS index + Context Profile
-> LiteContext CLI / MCP stdio
-> cited local context for an AI client@litecontext/core— contracts, profiles, SQLite/FTS index, retrieval.@litecontext/connectors— portable source connectors.@litecontext/cli— source lifecycle commands, diagnostics, and MCP server.
Development
The transition repository currently verifies both the TypeScript baseline and the Go read-only engine. It requires Go 1.24.1, Node 24.13–24.x, and pnpm 10.
pnpm install
pnpm check
pnpm audit:securitypnpm check runs the shared golden fixture tests for both implementations.
See CONTRIBUTING.md, SECURITY.md, third-party notices, and the release guide.
License
This server cannot be deployed
Maintenance
Related MCP Connectors
Serve a folder of Markdown notes as an MCP server: hybrid search, reading, and sourced answers.
Agent-native MCP server over the public saagarpatel.dev corpus. Read-only, stateless.
Personal knowledge base MCP server with semantic search, auto-categorization, metadata extraction
Related MCP Servers
- AlicenseAqualityDmaintenanceA local-first MCP server that exposes personal notes and files as unified semantic context for AI agents via vector search and file monitoring.6MIT
- AlicenseAqualityDmaintenanceLocal-first RAG indexing and semantic search MCP server. Enables document retrieval and context-aware queries using local embedding models.36MIT
- AlicenseNot gradedqualityDmaintenanceA local-first MCP server that turns a .context/ folder of markdown files into a searchable knowledge layer for AI coding agents.122MIT
- AlicenseNot gradedqualityAmaintenanceLocal-first MCP server for retrieval over markdown wikilink vaults, offering hybrid vector+lexical search, note reading, neighbor expansion, and recent activity tracking with fully local embeddings and no network egress.MIT