harness-document-project

Documenting a brownfield project for AI coding agents involves scanning the existing codebase to generate structured artifacts like source tree analysis and architecture docs, ensuring agents understand constraints before modifying code.

Providing AI context for an existing repository requires converting missing project context into structured documentation artifacts, utilizing scan-depth options and resumable state tracking to capture architecture and risk-aware outputs.

Yes, you can resume an interrupted project scanning workflow because the process maintains an on-disk state report within project_knowledge, allowing interrupted deep-dive analysis runs to safely resume.

Yes, the brownfield documentation workflow supports YAML-defined behavior, enabling customization resolution, config loading, and persistent-facts ingestion to structure the project scanning process.

Available scan-depth options for generating project architecture docs include quick, deep, and exhaustive modes, allowing you to control the scan level and produce exhaustive deep-dive analysis or rapid initial scans.

Yes, you need python3 installed as a dependency to run the document-project workflow, which processes the codebase and outputs structured documentation artifacts for AI agents.