project-knowledge-extraction

What problem does it solve?

This Skill prevents teams from repeatedly rediscovering the same project facts by turning scattered project evidence into durable, evidence-backed agent context.

Core Features & Use Cases

• Evidence inventory & source discovery: Scans code, docs, issues, incidents, reports, tests, scripts, and screenshots to find relevant material.
• Stable knowledge extraction & classification: Extracts durable facts, vocabulary, boundaries, and recurring failure modes, then classifies them into skills, ADRs, context docs, glossaries, runbooks, or memory.
• Grounding, freshness, and drift awareness: Attaches truth sources and verification metadata so the knowledge can be checked for drift over time.
• Graph linking for future reuse: Links new artifacts into the context graph so later agents can find and apply the knowledge.

Quick Start

Use the project-knowledge-extraction skill to mine this repository’s code and documentation into grounded, reusable agent context artifacts.