Capture what to keep, stop, and start after an initiative, sprint, or kill. Closes the PDLC cycle by turning experience into durable knowledge.
Scanned 10/6/2026
npx -y skills add tomzx/agents --skill run-retrospective --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Run Retrospective?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/tomzx-run-retrospective)More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.
---
name: run-retrospective
description: Capture what to keep, stop, and start after an initiative, sprint, or kill. Closes the PDLC cycle by turning experience into durable knowledge.
argument-hint: "[initiative-id or scope]"
---
# Run Retrospective
Turns a completed initiative, a sprint, or a killed bet into organizational knowledge. Produces a learning record in `.pdlc/learnings/`. A kill without a retrospective wastes the effort — the point of killing is to learn cheaper than building would have taught.
## Prerequisites
- Apply the shared PDLC conventions in `skills/pdlc/references/shared.md`.
- A concluded initiative (shipped, iterated, or killed), or a sprint boundary.
## Steps
1. Frame the scope: which initiative or period, and what outcome occurred (shipped, iterated, killed).
2. Capture what to **keep** (what worked and should be repeated), what to **stop** (what to never do again), and what to **start** (new actions to try).
3. For each item, name the root cause, not just the symptom. "We were slow" is a symptom; "discovery and build ran in series with no handoff contract" is a root cause.
4. Tie each learning back to the artifact chain: which assumption, decision, or gate should have caught this earlier?
5. Convert the strongest start/stop items into concrete next actions with owners (feed the roadmap or a new initiative).
6. Write the record to `.pdlc/learnings/N-<slug>.md`.
## Output Format
Use the template at `skills/pdlc/templates/learning.md`. Keep/stop/start sections, each item with root cause and (where relevant) a traceable artifact link.
## Outcome
If `$OUTCOME_YAML` is set, emit `verdict: captured`.
## Completion Checklist
- [ ] Keep / stop / start each populated
- [ ] Each item has a root cause, not just a symptom
- [ ] Learnings traced back to the artifact chain where possible
- [ ] Strongest items converted to owned actions
## Next Step
Feed owned actions into `build-roadmap` or a new `/pdlc discover` cycle.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!