Summary

stevesolun/ctx is a real-time recommendation engine for Claude Code, addressing the discovery and management of its extensive skill/agent ecosystem. It intelligently suggests relevant tools on the fly, optimizes context budgets, and mitigates "skill rot" by flagging unused components. The core benefit is enhanced developer productivity and more efficient LLM interactions via dynamic tool selection.

How It Works

The engine monitors code development, traversing a knowledge graph of 1,700+ skills, 400+ agents, and 10,000+ MCP servers (13,041 nodes, 847K edges). This blended graph integrates semantic, tag, and slug-token data. ctx hooks into Claude Code events, recommending and installing appropriate tool bundles based on current context. Its persistent LLM wiki memory learns across sessions, ensuring optimal tool access without token waste.

Quick Start & Requirements

• Installation: pip install claude-ctx
• Setup: ctx-init --hooks for one-shot directory/hook setup.
• Optional Dependencies: pip install "claude-ctx[embeddings]" (semantic backend), pip install "claude-ctx[dev]" (test toolchain).
• Pre-built Knowledge Graph: Download graph/wiki-graph.tar.gz from GitHub releases; extract to ~/.claude/skill-wiki/ using tar xzf graph/wiki-graph.tar.gz -C ~/.claude/skill-wiki/ (use --force-local on Windows/Git-Bash).
• Documentation: https://stevesolun.github.io/ctx/

Highlighted Details

• Knowledge Graph: Extensive graph with 13,041 nodes and 847,000 edges.
• Real-time Contextualization: Dynamically recommends and installs tools based on active development.
• Context Budget Optimization: Targets loading only 10-15 skills per session.
• Local Dashboard: ctx-monitor serve offers an interactive UI for skill management, graph visualization, LLM wiki browsing, and session activity monitoring.

Maintenance & Community

No specific details regarding maintainers, community channels (e.g., Discord, Slack), or project roadmap were present in the provided README excerpt.

Licensing & Compatibility

Released under the MIT License, which is permissive for commercial use and integration into closed-source projects.

Limitations & Caveats

The provided README excerpt does not detail specific limitations, known bugs, or alpha/beta status, focusing instead on the problem domain and solution capabilities.