retrospective

To capture lessons learned after task completion, run a structured retrospective workflow that reflects on what worked and what failed, then proposes actionable updates to your AI-assisted development instructions and knowledge files.

Improving future AI workflows requires post-task reflection to document debugging insights, classify the outcomes, and route recommended updates to instructions, knowledge files, or issues for continuous improvement.

A post-task review workflow handles knowledge management by enforcing a multi-step process that captures task reflections, preserves knowledge for repeatable improvements, and routes classified updates to relevant project files.

Yes, you can use a retrospective workflow to update instructions and issues after non-trivial software tasks, as it classifies lessons learned and explicitly recommends routing those updates to your knowledge files or issue tracker.

You should run a post-task retrospective immediately after completing non-trivial software engineering work, such as feature development or debugging, to ensure task context is preserved and routed into knowledge files or instructions.

No specific dependencies are required to run a structured retrospective for AI-assisted development, as the workflow operates independently to classify lessons learned and recommend updates to your existing knowledge files and instructions.