skill-evolution-advisor

To log recurring workflow friction points during research sessions, use the Skill to capture lightweight user preferences and recurring issues directly within your research workspace. It tracks these system defects to facilitate iterative improvement.

Yes, you can use the skill-evolution-advisor to remember your formatting preferences. It captures lightweight research learnings and user habits, allowing the system to adapt to your specific research style and avoid repeated corrections.

Yes, this Skill requires manual confirmation gates before changing agent behavior. This structured diagnostic logging process ensures privacy and deterministic evolution of your workflows rather than automatic, unverified system updates.

To generate actionable improvement prompts when research workflows fail, the Skill operates within your research workspace to track friction points and record defects. It then produces structured retrospective prompts for workflow optimization.

The limitation of using automated memory capture for workflow diagnostics is that it requires structured diagnostic logging and manual confirmation gates. This ensures deterministic evolution of agent behavior, meaning changes are not applied automatically without verification.