sprint-report

name	sprint-report
description	Generate a sprint status report for the current phase. Default is --table.

Sprint Report Skill

Build fenced JSON and pipe to the Jinja2 template. mode controls table vs detailed.

Usage

/sprint-report [--table | --detailed]

Default: --table

---

Data Source

Always use gh pr list first - single call, returns all open PRs with CI and merge state:

gh pr list --state open --json number,title,headRefName,baseRefName,isDraft,statusCheckRollup

This is faster and sufficient for populating sprint_rows and integration_row. Only drill into individual gh run view calls if you need failure details for a specific job.

Dogfooding rule: If gh pr list output is missing information needed to fill the report (e.g., no per-job failure detail, no QA state, truncated CI summary), file a GitHub issue describing what field or format change would make it sufficient, then improve the command. Do not silently work around gaps with extra gh CLI calls - surface them as product issues.

Render Command

The template path is relative - must run from the main repo root (not a worktree).

cd "${CLAUDE_PROJECT_DIR:-$(git worktree list | head -1 | awk '{print $1}')}"
echo '<json>' > /tmp/sprint-report.json
sc-compose render .claude/skills/sprint-report/report.md.j2 --var-file /tmp/sprint-report.json

--table (default)

{
  "mode": "table",
  "sprint_rows": "| AK.1 | ✅ | ✅ | 🏁 | #621 |\n| AK.2 | ✅ | ✅ | 🌀 | #622 |",
  "integration_row": "| **integrate** | | — | 🌀 | — |"
}

--detailed

{
  "mode": "detailed",
  "sprint_rows": "Sprint: AK.1  Contract reconciliation\nPR: #621\nQA: PASS ✓ (iter 3)\nCI: Merged to integrate/phase-AK ✓\n────────────────────────────────────────\nSprint: AK.2  OTel core\nPR: #622\nQA: PASS ✓\nCI: Running (1 pending)",
  "integration_row": "Integration: integrate/phase-AK → develop\nCI: Running — pending AK.4 + AK.5"
}

Icon Reference

State	DEV	QA	CI
Assigned	📥	📥
In progress	🌀	🌀	🌀
Done/Pass	✅	✅	✅
Findings	🚩	🚩
Fixing	🔨
Blocked			🚧
Fail			❌
Merged			🏁
Ready to merge			🚀