improvement-loop

You refine AI pipeline components by systematically updating skills, agents, and hooks after issue resolution. This targets agent anti-patterns and script error handling to prevent the recurrence of identified problems in your workflow.

The pipeline skill enhancement process follows a strict five-step cycle: capture, classify, change, verify, and commit. This systematically updates skills, agents, and hooks after issue resolution to prevent problem recurrence.

To add anti-pattern rules after a bug is resolved, execute the capture, classify, change, verify, and commit cycle. You classify the error as an anti-pattern, change the agent's definition to include the new rule, and verify the update prevents future mistakes.

Yes, you can proactively optimize agent workflows by identifying and suggesting improvements for recurring inefficiencies observed during operation. This targets the refinement of skill guidance and hook logic before actual errors disrupt the pipeline.

Workflow optimization refines four main AI pipeline components: skills, agents, hooks, and scripts. It specifically targets agent anti-patterns, skill guidance, hook logic, and script error handling to ensure a more robust workflow over time.

No specific dependencies are required to run pipeline improvement scripts. The workflow operates independently to refine agent anti-patterns and skill guidance using its internal five-step capture, classify, change, verify, and commit cycle.