self-improve

Generating actionable todos from session logs requires analyzing session artifacts like errors and subagent summaries to identify concrete improvements. The retrospective process surfaces optimization opportunities across agent configuration, tests, docs, and code, prioritizing findings with suggested owners and rationale.

An end-of-session retrospective for agent configuration and code improvement is an automated review of session artifacts to specify recommended changes. It analyzes errors, logs, and tool outputs to produce ready-to-execute todos with measurable next steps and required scripts for engineering teams.

Yes, you can use session artifacts like tool outputs and subagent summaries for continuous improvement planning. Analyzing these artifacts identifies concrete optimization opportunities across documentation, tests, and code, transforming findings into actionable todos for product and engineering teams.

Prioritizing improvements from a post-release retrospective involves analyzing session artifacts to surface concrete optimization opportunities. The approach suggests owners, provides rationale for recommended changes, and generates ready-to-execute todos with measurable next steps for doc and engineering teams.

No dependencies are required to analyze session errors and generate documentation improvements. The process operates standalone, directly reviewing session artifacts like logs and tool outputs to identify concrete improvements and produce actionable todos with recommended changes and measurable next steps.