Project Plan

Current implementation status

The project already includes a working foundation that covers the main Auto Doc workflow from repository analysis to documentation publishing.

Completed and active capabilities

1. A FastAPI service has been implemented to coordinate repository analysis, documentation generation, and publishing workflows.

2. Repository access has been integrated for both GitHub and GitLab.

3. The system can scan Python, JavaScript, TypeScript, and MATLAB source files (.m/.matlab) and assess documentation coverage across supported source code blocks.

4. AI-powered docstring suggestion generation is available for missing documentation through OpenAI, Codex CLI, or Claude CLI, with support for reuse of previously generated suggestions.

5. A Sphinx documentation scaffold is generated automatically for reviewed repositories.

6. AutoAPI integration is supported through the autoapi_include/ workflow, including filtering of risky files and skipped-file reporting during build fallback.

7. The project supports a review-first publishing flow in which documentation can be built and published to GitHub Pages.

8. Runtime logs and analysis artifacts are stored to support traceability and review of each documentation run.

Near-term refinement work

1. Improve the generated documentation templates so project pages are more structured, easier to review, and more visually consistent.

2. Strengthen documentation quality checks and evaluation across generated outputs.

3. Refine the publishing and scaffold workflow to reduce friction for teams adopting the system in real repositories.

Future development

1. Introduce Retrieval-Augmented Generation (RAG) so docstring and documentation suggestions can be grounded in repository-specific context, existing documentation, architectural notes, and previously approved project knowledge.

2. Extend the platform toward audit-ready documentation by generating more structured evidence trails, change histories, traceable outputs, and documentation artifacts suitable for regulated or compliance-aware environments.

3. Expand support for stronger documentation intelligence, including richer contextual recommendations, improved summarization, and more reusable documentation outputs across project lifecycles.

Validation approach

1. Test the workflow against representative repositories from supported languages and repository providers.

2. Review the usefulness, clarity, and correctness of generated docstring suggestions.

3. Assess the quality of generated Sphinx pages, published HTML output, and supporting logs and artifacts.

4. Use developer feedback to guide improvements in usability, trust, and documentation completeness.

Milestones

Milestone	Outcome
M1	Core FastAPI service and repository analysis implemented
M2	Docstring generation and Sphinx scaffold workflow active
M3	Review-first publishing and reporting workflow stabilized
M4	RAG and audit-ready documentation features planned