synthesize-tech-docs

To generate technical documentation aligned with your PRD and codebase, synthesize your product requirements, design docs, and production code together. This process creates markdown artifacts like READMEs and architecture overviews while explicitly flagging discrepancies between planned intent and built features.

Documenting discrepancies between design docs and actual implementation involves triple-source synthesis. By comparing the product requirements and design documentation directly against the production codebase, you can explicitly surface mismatches in your engineering technical documentation before they cause downstream issues.

Yes, you can generate an API reference and architecture overview from a local codebase and a design doc. By requiring all three source inputs—the PRD, design doc, and codebase—the synthesis process ensures the resulting markdown artifacts capture business intent, UX flows, and actual implementation details accurately.

Yes, both a PRD and a design doc are required to create engineering documentation for a new feature. The tool enforces that all three source inputs are present, including the production codebase, to avoid generating speculative or incomplete documentation that misses critical business intent or UX context.

To publish generated technical documentation to Confluence, use the optional Confluence publishing feature after synthesizing your PRD, design doc, and codebase. This produces aligned markdown artifacts such as READMEs and API references, which can then be exported directly to your Confluence workspace.

New feature documentation has gaps between user experience design and actual code implementation when generated without synthesizing all three sources. By cross-referencing the PRD, design doc, and production codebase together, the triple-source synthesis explicitly surfaces mismatches to address these discrepancies before they cause miscommunication.