update-docs

You can automate updating project documentation before a pull request by delegating the task to a sub-agent. This offloads large file processing, minimizing the main agent's context consumption while ensuring CLAUDE.md and implementation-status.md reflect recent changes.

Agent context gets bloated when updating CLAUDE.md because processing large documentation files directly consumes significant tokens. Delegating this documentation update task to a sub-agent prevents context bloat by offloading the heavy file processing from the main agent.

The best way to update implementation status docs after a feature is using a sub-agent delegation workflow. It requires summarizing implementation details, steering directory paths, and optionally providing test coverage changes to accurately update docs/implementation-status.md.

Yes, you can delegate documentation updates to minimize main agent context. This workflow offloads the processing of large files like CLAUDE.md to a sub-agent, ensuring the main agent's context remains free for core development tasks.

To update project documentation via a sub-agent, you need to provide summaries of implementation details, steering directory paths, and optionally test coverage changes. This ensures the sub-agent accurately updates CLAUDE.md and implementation-status.md.