improve-skill

To generate improvements for an AI workflow, analyze the most recently executed skill to identify gaps and produce targeted recommendations for skill.md updates. This refines future performance through structured post-task reflection.

The best way to refine skill documentation is to analyze post-execution feedback to surface clarity issues and edge cases, then apply precise edits to the target skill.md and workflow steps for iterative improvement.

Yes, you can use post-task reflection to update any skill without external tools. The process requires only the current interaction to analyze the just-used skill and generate concrete documentation updates.

To identify gaps in a recently executed skill, evaluate its post-execution performance to surface clarity issues, edge cases, and output quality problems. This analysis guides targeted changes for the next activation.

You need iterative refinement when a previously executed skill shows output quality issues or unhandled edge cases. It provides a structured plan to apply concrete improvements on the skill's next activation.