project-completion-knowledge-extraction

Distribute completed sandbox work using a three-tier documentation architecture that separates lightweight production READMEs, comprehensive knowledge bases, and agent pattern indexes. It requires separate pull requests for production repositories and the knowledge base.

A three-tier documentation architecture separates project handoffs into lightweight repository READMEs, centralized knowledge bases with deployment and operations details, and agent pattern indexes. This structure ensures fast onboarding and robust learning across repositories.

Extract knowledge from completed sandboxes by enforcing a knowledge extraction workflow that captures architecture, deployment, and operations documentation in a centralized knowledge base. It also indexes reusable agent patterns before archiving the sandbox environment.

Independent pull requests prevent mixing changes across production repos and knowledge bases, preserving clear ownership and traceability. This workflow ensures code, documentation, and patterns remain cleanly separated during sandbox distribution.

No external dependencies are required to distribute sandbox work to proper repositories. The workflow relies on standard cross-repo pull request flows to move completed code, generate lightweight READMEs, and populate a centralized knowledge base.