productionos-session-validate

End-of-session learning reports are generated by reading session data and extracting patterns with metaclaw-learner to produce concise summaries. The Skill applies to Codex and Claude workflows across quick, standard, or deep modes to capture session metrics and output optimization hypotheses.

Extracting recurring patterns from AI session metrics uses metaclaw-learner to analyze end-of-session artifacts and identify lessons. This mechanism updates instincts and generates optimization hypotheses, guiding iterative improvements for future ProductionOS runs.

Session validation supports quick, standard, or deep modes for ProductionOS workflows. You can apply these varying depths to both Codex and Claude end-of-session reviews depending on the level of detail required for your session report.

End-of-session validation works directly with both Codex and Claude end-of-session reviews. It reads the session data from these workflows, extracts patterns using metaclaw-learner, and writes a session report with optional optimization hypotheses for the next run.

Optimization hypotheses guide iterative improvements by capturing lessons from previous session metrics. Generating these hypotheses at the end of a session updates instincts and provides actionable directions to boost the performance of future runs.

Session analytics with metaclaw-learner focus specifically on end-of-session artifacts, metrics, and pattern extraction within ProductionOS workflows. It is designed to summarize concluded sessions and generate next-run hypotheses rather than monitor live or ongoing sessions.