ctx2skill

What problem does it solve?

Ctx2Skill automatically discovers, refines, and selects context-specific skills from complex documents using a self-evolving multi-agent loop, eliminating the need for manual annotation or external feedback.

Core Features & Use Cases

• Self-play loop with five agent roles (Challenger, Reasoner, Judge, Proposer, Generator) to extract reusable context knowledge.
• Cross-Time Replay to identify robust skills that generalize across tasks and time.
• Markdown-form output that can be prepended to system prompts during inference, enabling zero or minimal downtime deployment.
• Command-line and Python API access for easy integration into existing LLM workflows.

Quick Start

Run the module with your context text and an OpenAI-compatible API, then feed the produced skill set into your next inference task.