Session Review

The session-end review should be short.

It exists to capture actionable framework friction while the evidence is still fresh, not to produce a long essay.

Questions

Ask:

• what slowed progress?
• what was confusing?
• what repeated manual effort?
• what should have been easier?
• what should ORCA Framework have already known or handled?
• did ORCA Framework create unnecessary setup work or too many decisions?
• would one small user-facing coaching hint have reduced friction next time?
• what feature, doc, guide, command, or automation was missing?
• was this a one-off or a reusable framework problem?
• did the user seem frustrated based on observable evidence?
• did the output seem repetitive, generic, templated, or shallow?
• did ORCA Framework keep doing the wrong thing after correction?
• did it recover well from errors?
• did it create concrete value or mostly polished motion?

Output Shape

Use templates/session-improvement-review.md.

The review should fit on one screen in most cases. Include a short quality-signal summary, not a second long retrospective.

Trigger Rule

Prefer the review after meaningful sessions, blocked sessions, or surprising sessions.

Do not force it after every trivial success.