crow-pet-self-improvement

Automating task feedback analysis involves parsing corrections, approvals, and failures to generate structured memory and policy updates. This Skill processes experiential feedback files to propose behavioral adjustments, reducing the need for manual retraining and human oversight.

Updating agent memory from failure patterns requires parsing feedback files to identify recurring issues. This Skill reflects on task outcomes, converts experiential feedback into reusable lessons, and proposes updates to rubrics and policies for improved future autonomy.

Structured reflection for task learning is the process of analyzing task outcomes to extract reusable lessons. This Skill automates that reflection by evaluating corrections, approvals, and failures to update behavior patterns and improve ongoing task execution without manual intervention.

Applying task feedback to update policies without manual retraining is achieved by loading feedback files like `memory/approvals.md`. This Skill parses observed patterns from corrections and failures, proposing direct updates to memory and policies to enhance future performance.

Yes, this feedback parsing approach works with approval and correction files by design. It handles feedback from different sources including corrections, approvals, and failures, converting them into structured memory updates and policy adjustments for autonomous task execution.

You should avoid automating memory updates when task feedback is unstructured or lacks clear patterns. This Skill requires parsing feedback files and proposing updates based on observed patterns, so highly ambiguous or isolated corrections may not yield reliable policy adjustments.