Skip to content

Architecture

The project keeps a small pipeline:

  1. RepositoryScanner reads useful text files from a local repository.
  2. PythonAnalyzer, StructureAnalyzer and DependencyAnalyzer produce typed facts.
  3. RepositoryRAG delegates indexing and retrieval to python-rag-framework.
  4. DocumentationAgent delegates prompt execution to python-agent-runtime.
  5. Markdown generators build focused documentation files.
  6. OutputWriter writes generated files to disk.

The default generation path is deterministic so tests and examples run without paid LLM APIs.

Hosted OpenAI, Anthropic and Gemini adapters are optional. They receive a deterministic draft as grounding, then rewrite it through the selected provider.