lean-ctx
Active·★ 3.9k·MIT·Updated 2026-10-09
★ Hidden Gem★ Dev Tooling★ Memory & Context
lean-ctx is an open-source MCP server designed to drastically reduce LLM token costs for AI coding tools like Claude Code, Cursor, and Copilot. It achieves this by transparently compressing CLI output and file context, offering up to 99% savings and making AI coding faster, cheaper, and more effective without changing workflows.
#agentic-coding#ai#ai-coding#claude-code#context-engineering#copilot#cursor#developer-tools
01
Features
01Transparent shell hook for CLI output compression
02Context server with 28 intelligent tools
03One-command integration for AI coding agents
04Adaptive LLM communication with Cognitive Efficiency Protocol
05Cross-session memory for context continuity
02
Compatibility
Cursor
Auto-configured
Verified via docs
Claude Code
Auto-configured
Verified via docs
GitHub Copilot
Auto-configured
Verified via docs
VS Code
Auto-configured
Verified via docs
JetBrains IDEs
Auto-configured
Verified via docs
03
Quick start
1
$ curl -fsSL https://leanctx.com/install.sh | sh
2
$ lean-ctx setup
3
$ lean-ctx doctor
04
Use cases
↳Reducing token costs for Claude Code, Cursor, and Copilot
↳Optimizing MCP server performance
↳Saving $30-100+ per month on AI coding API costs
05
Alternatives
Comments
Log in to leave a comment
No comments yet. Be the first!