Artifact: retrospective

A human-facing learning report produced at the close of a work period: what went well, what didn’t, and the lessons/action items future iterations should surface. gstack’s gstack-retro generates a team-aware weekly form — per-person breakdowns, shipping streaks, test-health trends, growth opportunities (/retro global spans all projects and AI tools). BMAD’s bmad-retrospective generates the post-epic Party-Mode form — a facilitated persona roundtable extracting lessons and action items.

The retrospective is the team-process flavor of the stage-learn stage — the human-facing counterpart to the machine-consumable artifact-solution-doc (Compound Engineering’s agent-grounding corpus, and gstack’s own gstack-learn). Same generic activity (harvest lessons → fold into a durable form that improves future work); different consumer (the human team vs future agents).

Produced by (backlinks)

• gstack-retro — team-aware weekly retrospective with per-person breakdowns and streaks.
• bmad-retrospective — post-epic Party-Mode retrospective extracting lessons + action items.

See Also

• artifact-solution-doc — the agent-facing learning artifact sibling (ce-compound, gstack-learn).
• pattern-knowledge-compounding — the technique both learning artifacts realize.
• stage-learn — the stage this artifact is the (team-process) output of.