run-retro — Claude Code command: what it does, how to install
run-retro orchestrates renewal cycle retrospectives to analyze ARR shifts, review churn patterns, and generate action trackers.
Apache-2.0 ★ 450 updated 2026-04-03
.claude/commands/run-retro.mdWhat run-retro does
The command structures post-renewal reviews across five steps: compiling deal metrics, grouping outcome patterns, extracting takeaways, assigning tasks, and creating presentation slides. It supports filters by segment (enterprise, growth, scale), review window, and grouping dimensions such as CS owner or geography.
Execution logs a .claude/plans/plan-<timestamp>.json file documenting dependencies across CS Ops, Product, and Finance. The workflow invokes agents like renewal-director and cs-ops-partner alongside the renewal-playbooks and deal-desk skills. If data pulls fail, the execution allows up to 2 retries before escalating.
Unresolved blockers recurring across two consecutive retrospectives trigger an escalation directly to CRO and Product leadership.
What it helps with
- run structured post-renewal retrospectives
- analyze churn reasons and ARR deltas across customer tiers
- generate action trackers with owners and due dates
- store orchestration plans under
.claude/plans
Example prompts for the agent
- /renewal-orchestration:run-retro --segment enterprise --window quarter --include-churn true
- /renewal-orchestration:run-retro --window rolling-90d --depth workshop
- /renewal-orchestration:run-retro --segment growth --dimensions cs-owner --include-churn false
How to install run-retro
Manually: put the files in place
.claude/commands/run-retro.mdCall it with /run-retro. For all projects use ~/.claude/commands/
Manually: put the files in place
~/.codex/prompts/run-retro.mdCall it with /prompts:run-retro. Codex marks this format deprecated in favour of skills
Manually: put the files in place
.github/prompts/run-retro.prompt.mdManually: put the files in place
.gemini/commands/run-retro.tomlGemini stores commands as TOML: the prompt text goes into the prompt field
Questions about run-retro
How to install run-retro in Claude Code?
Place the command definition file at .claude/commands/run-retro.md.
What does run-retro do?
It runs an end-to-end post-renewal retrospective, compiling deal health and ARR trends, identifying churn drivers, and documenting action items.
What parameters does run-retro support?
It accepts segment (enterprise, growth, scale), window (month, quarter, rolling-90d), include-churn, dimensions (such as cs-owner or geo), and depth (summary or workshop).
Source text
run-retro.md · Apache-2.0
--- name: run-retro description: Facilitates structured retrospectives after renewal cycles to capture insights, update playbooks, and improve forecasting. usage: /renewal-orchestration:run-retro --segment enterprise --window quarter --include-churn true --- # Command: run-retro ## Inputs - **segment** – optional filter (enterprise, growth, scale) for the retro scope. - **window** – time range (month, quarter, rolling-90d). - **include-churn** – true/false toggle to include churned accounts alongside renewals. - **dimensions** – optional grouping (cs-owner, product, geo, persona). - **depth** – summary | workshop (adds facilitation guides + templates). ### GTM Agents Pattern & Plan Checklist > Derived from GTM Agents orchestrator blueprint @puerto/plugins/orchestrator/README.md#112-325. - **Pattern selection**: Retros typically follow **pipeline** (data → analysis → insights → action planning → documentation). If insight synthesis + action planning can run in parallel (workshop mode), log a **diamond** segment with merge gate. - **Plan schema**: Save `.claude/plans/plan-<timestamp>.json` with objective, scope, task IDs, parallel groups, dependencies (CS Ops, Product, Finance), error handling, and success metrics (closed-loop actions %, renewal uplift, blocker removal). - **Tool hooks**: Reference `docs/gtm-essentials.md` stack—Serena for pulling historical plan diffs, Context7 for product/support notes, Sequential Thinking for facilitation prompts, Playwright for verifying embedded dashboards if used. - **Guardrails**: Default retry limit = 2 for data pulls or workshop sync failures; escalation ladder = Renewal Director → CS Ops Partner → Exec Sponsor. - **Review**: Use `docs/usage-guide.md#orchestration-best-practices-puerto-parity` before execution to confirm facilitation resources, dependencies, deliverables. ## Workflow 1. **Data Compilation** – aggregate deal outcomes, ARR deltas, health trends, product usage shifts, and qualitative notes. 2. **Pattern Analysis** – bucket learnings by reason codes, plays applied, executive involvement, and pricing. 3. **Insight Synthesis** – highlight leading indicators, blockers, and emerging opportunities. 4. **Action Planning** – assign remediation owners, timeline, and metrics for follow-up workstreams. 5. **Documentation** – publish retro deck + database entries for ongoing knowledge base. ## Outputs - Retro report summarizing metrics, insights, and improvement themes. - Action tracker with owners, due dates, and success criteria. - Template slides + transcripts for sharing across CS, product, finance leadership. - Plan JSON entry stored/updated in `.claude/plans` for audit trail. ## Agent/Skill Invocations - `renewal-director` – sponsors session and finalizes recommendations. - `cs-ops-partner` – compiles data and action tracker. - `renewal-playbooks` skill – integrates learnings into new plays. - `deal-desk` skill – informs pricing/process adjustments. ## GTM Agents Safeguards - **Fallback agents**: document substitutes (e.g., Exec Sponsor covering Renewal Director) when facilitators unavailable. - **Escalation triggers**: if critical blockers repeat across two retros or action items stall beyond SLA, escalate to CRO + Product leadership per GTM Agents rip-cord. - **Plan maintenance**: update plan JSON/change log when retro scope, owners, or follow-up workstreams change to maintain GTM Agents audit parity. ---
© gtmagents · Apache-2.0 · gtmagents/gtm-agents