review-docs

To review technical documentation for accuracy and style issues, you can run parallel style and content evaluations that surface gaps across developer guides, READMEs, and API docs. This process identifies clarity problems and maintains a state file to track findings across multiple review rounds.

Evaluating API documentation quality is best handled by running concurrent style and content evaluators that identify gaps in clarity and accuracy. The synthesis output summarizes findings and prompts you for action, enabling iterative improvements across multiple review rounds.

Yes, you can perform multi-round documentation reviews to verify fixes using a maintained scratchpad state file. This state tracking mechanism records identified issues across rounds, allowing you to confirm whether previous style and accuracy gaps have been resolved.

You can improve README and developer guide clarity by initiating a parallel evaluation loop that separately checks style and content. The evaluation produces a synthesis of findings with next-step prompts, guiding you to iteratively apply content improvements until completion.

Yes, the documentation review process works without external dependencies, requiring only your target technical content files as input. It operates autonomously to orchestrate parallel style and content evaluations, track issues in a state file, and output a synthesis of findings.

Automated documentation evaluation requires user action to apply fixes, as it only surfaces style, clarity, and accuracy issues rather than directly modifying files. It relies on iterative user prompts to improve, complete, or finish the review loop based on the synthesized findings.