sprint-start
Skill9 starsChanged 3 months agoMITView on GitHub
Pre-sprint checks, feature branch creation, and sprint planning. Ensures clean state, defines sprint goal and scope, creates sprint spec document.
- Commits and pushes
In this repository56 other files it gives its agents.
- .claude/agents/CLAUDE.mdCLAUDE.md
- .claude/hooks/CLAUDE.mdCLAUDE.md
- CLAUDE.mdCLAUDE.md
- .claude/prompts/CLAUDE.mdCLAUDE.md
- .claude/rules/CLAUDE.mdCLAUDE.md
- .claude/skills/CLAUDE.mdCLAUDE.md
- docs/CLAUDE.mdCLAUDE.md
- vision/CLAUDE.mdCLAUDE.md
- llms.txtllms.txt
- scaffold/llms.txtllms.txt
- architecture-checkSkill
- backlog-reviewSkill
- bootstrapSkill
- brainstormSkill
- buildSkill
- claude-sense-checkSkill
- code-qualitySkill
- commitSkill
- continueSkill
- custom-hooksSkill
- dashboardSkill
- debug-sessionSkill
- deploySkill
- discoverSkill
- doctorSkill
- fix-issueSkill
- framework-upgradeSkill
- handoffSkill
- help-meSkill
- ideateSkill
- lanesSkill
- manual-testSkill
- merge-downSkill
- merge-upSkill
- optimizeSkill
- parallel-workSkill
- performance-checkSkill
- phase-reviewSkill
- pr-statusSkill
- quality-checkSkill
- quickstartSkill
- refine-loopSkill
- researchSkill
- retrospectiveSkill
- security-auditSkill
- skill-createSkill
- skill-evalSkill
- sprint-endSkill
- story-cycleSkill
- testing-cycleSkill
- test-validatorSkill
- UAT-cycleSkill
- undo-workSkill
- uninstallSkill
- weekly-maintenanceSkill
- wrap-upSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
27 rules. About 0 are common, 19 are distinctive, and 8 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈19
Distinctive · estimate
This author's own choices. Worth a look.
8
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 26
Read the Profile: line from CLAUDE.md to determine the active project profile.
- Distinctive: line 33
Strict mode: All steps mandatory. Metrics and debt checks must produce actionable output (not just "all clear").
- Distinctive: line 38
Verify the workspace is ready for new work:
- Very specific: line 48
Check if any are approved → merge them: gh pr merge --squash --delete-branch
- Distinctive: line 49
If awaiting review → inform user and ask whether to proceed or wait
- Distinctive: line 60
Do NOT proceed with dirty working tree
- Distinctive: line 64
Read the Default branch from CLAUDE.md's Git Workflow section.
- Distinctive: line 87
If the default branch was detected at runtime (not in CLAUDE.md), persist it by updating CLAUDE.md's Git Workflow section.
- Distinctive: line 91
Read CLAUDE.md Commands section to find the project's test command.
- Distinctive: line 93
If test command exists: Run it. If tests fail on the default branch, stop and alert user — default branch should always be green.
- Very specific: line 98
Read docs/progress.md → ## Metrics table.
- Distinctive: line 113
If all metrics are 🟢 with stable trends, or if the Metrics table has no data yet, skip this section silently.
- Distinctive: line 131
If Critical items exist, present them as P0 stories alongside backlog stories in the step 3 story table.
- Distinctive: line 140
Create a new branch from the default branch:
- Distinctive: line 148
Create an isolated worktree for parallel development:
- Distinctive: line 174
After branch creation, ask the user if they want to create a draft PR for early CI feedback:
- Distinctive: line 183
If user declines, skip — the branch is pushed and PR created at sprint-end.
- Very specific: line 190
Read docs/reference/BACKLOG_INDEX.md and scan epic files for stories with status ready, grouped by priority:
- Very specific: line 200
If no stories have ready status, check for draft stories and suggest running /ideate to refine them, or /backlog-review to assess backlog …
- Distinctive: line 204
Ask the user to select stories for this sprint and define a sprint goal.
- Distinctive: line 229
Leave the Decisions, Notes, and Outcome sections empty — they're filled during and after the sprint.
- Very specific: line 233
Update docs/progress.md → ## Current Sprint section with:
- Very specific: line 261
Always show the actual first story ID from the sprint plan (e.g., /story-cycle E01-S01), not a placeholder.
- Very specific: line 271
Does not load story context (that's /story-cycle's job)
- Distinctive: line 272
Does not run analysis agents
- Distinctive: line 273
Does not update epic files or backlog
- Distinctive: line 274
Does not assume any particular project structure
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
sprint-start
Skill metrics: Emit a start event to the activity log:
echo "{\"type\":\"skill\",\"event\":\"start\",\"skill\":\"sprint-start\",\"ts\":\"$(date -u +%Y-%m-%dT%H:%M:%SZ)\"}" >> docs/sessions/.activity-log.jsonlStarting a new sprint.
Profile-Adaptive Behavior
Read the **Profile:** line from CLAUDE.md to determine the active project profile.
<IF condition="Profile is lean">
**Lean mode:** Skip steps 1.5 (Metrics Health Check), 1.6 (Debt Health Check), and step 3 (Sprint Planning ceremony — no sprint spec generation, no story selection, no PRD check). Perform only: 1a-1d (pre-flight checks), step 2 (create feature branch), and update `docs/progress.md` with branch name and sprint number.
</IF><IF condition="Profile is strict">
**Strict mode:** All steps mandatory. Metrics and debt checks must produce actionable output (not just "all clear"). Sprint spec generation is required with explicit quality gate plan section.
</IF>1. Pre-flight Checks
Verify the workspace is ready for new work:
1a. Check for open PRs
gh pr list --author @me --state openIf open PRs exist:
- Check if any are approved → merge them:
gh pr merge --squash --delete-branch - If awaiting review → inform user and ask whether to proceed or wait
1b. Verify clean working tree
git statusIf uncommitted changes exist:
- Warn user — they must commit, stash, or discard before proceeding
- Do NOT proceed with dirty working tree
1c. Ensure on default branch and up to date
Read the Default branch from CLAUDE.md's Git Workflow section. If not set, detect it at runtime:
# Read from CLAUDE.md first, fall back to detection
DEFAULT_BRANCH=$(grep -oP '^\- \*\*Default branch:\*\* \K\S+' CLAUDE.md 2>/dev/null)
if [ -z "$DEFAULT_BRANCH" ]; then
DEFAULT_BRANCH=$(git symbolic-ref refs/remotes/origin/HEAD 2>/dev/null | sed 's@^refs/remotes/origin/@@')
fi
if [ -z "$DEFAULT_BRANCH" ]; then
for branch in main master develop; do
if git show-ref --verify --quiet "refs/heads/$branch"; then
DEFAULT_BRANCH="$branch"
break
fi
done
figit checkout $DEFAULT_BRANCH
git pull origin $DEFAULT_BRANCHIf the default branch was detected at runtime (not in CLAUDE.md), persist it by updating CLAUDE.md's Git Workflow section.
1d. Verify tests pass on default branch
Read CLAUDE.md Commands section to find the project's test command.
- If test command exists: Run it. If tests fail on the default branch, stop and alert user — default branch should always be green.
- If NO test command configured: Skip this check. Note in output: "No test command configured — consider running /bootstrap to set up."
1.5. Metrics Health Check
Read docs/progress.md → ## Metrics table. If the table has data (not all "—"), check for quality signals that should influence sprint planning:
| Metric signal | Planning guidance |
| Cycle time ↑ or 🟡/🔴 | Fewer or smaller stories this sprint |
| Change failure rate 🟡/🔴 | Prioritize quality: include a bugfix or stabilization story |
| Code churn ratio 🟡/🔴 | Include a refactoring story targeting hotspot files (see sprint note for file list) |
| Test coverage Δ declining 3 sprints | Enforce strict TDD compliance for all stories |
| Sprint satisfaction declining 3 sprints | Investigate root cause before adding feature work |
| AI effectiveness declining | Review skill failure patterns; consider simpler story decomposition |
Present findings to the user before story selection in step 3, framed as recommendations not blockers:
"Sprint metrics check: Code churn ratio is 🟡 (0.14, target ≤0.15). Consider including a refactoring story targeting [hotspot files from sprint note]."
If all metrics are 🟢 with stable trends, or if the Metrics table has no data yet, skip this section silently.
1.6. Debt Health Check
<IF condition="docs/technical-debt.md exists and contains active items (not just template comments)">
Scan `docs/technical-debt.md` for debt that should influence sprint planning:- Active counts — read the header (
Active items: X). If 0, skip this section. - Sprint candidates — scan for items matching any of:
- Severity: Critical (must be addressed this sprint)
- Interest: Growing (compounding — gets worse each sprint it's deferred)
- Priority score ≥ 4.5 (high impact-to-effort ratio)
- Debt neglect check — scan the Resolved section for dates. If no items resolved in the last 2 sprints, flag: "No debt resolved in 2+ sprints — debt is accumulating without remediation."
Present findings before story selection in step 3, framed as recommendations:
"Debt register: [N] active items ([N] critical, [N] growing). Sprint candidates: TD-NNN [title], TD-NNN [title]. Consider including at least 1 debt remediation story."
If Critical items exist, present them as P0 stories alongside backlog stories in the step 3 story table. If Growing items exist, note them as recommended additions.
</IF>
2. Create Feature Branch
Determine the next sprint number by reading docs/progress.md and finding the highest sprint number, then adding 1.
Standard Mode (default)
Create a new branch from the default branch:
git checkout -b sprint-<number>Worktree Mode (when --worktree is in $ARGUMENTS or user requests it)
Create an isolated worktree for parallel development:
# Create worktree in sibling directory
git worktree add ../$(basename $(pwd))-sprint-<number> -b sprint-<number>Inform the user:
**Worktree created:** `../<project>-sprint-<number>`
To work in this worktree, open a new Claude Code instance in that directory.
Each worktree has its own branch and working tree, so you can work on multiple stories in parallel.
**Important:** When done, run `/sprint-end` from within the worktree. It will clean up after merge.Branch Naming
- Branch naming convention:
sprint-<number>(e.g.,sprint-001,sprint-002) - The number is always the next sequential sprint number
Optional: Draft PR for Early CI Feedback
<IF condition="CI is configured (.github/workflows/ detected) AND gh CLI is available">
After branch creation, ask the user if they want to create a draft PR for early CI feedback:git push -u origin sprint-<number>
gh pr create --draft --title "Sprint <N>: <sprint-goal>" --body "Work in progress — sprint branch for early CI feedback."Draft PRs enable CI to run on every push during the sprint, catching issues early. They block merging and suppress CODEOWNERS notifications until marked ready.
If user declines, skip — the branch is pushed and PR created at sprint-end.
</IF>
3. Sprint Planning
Show Upcoming Stories
Read docs/reference/BACKLOG_INDEX.md and scan epic files for stories with status ready, grouped by priority:
### Ready Stories (by priority)
| Priority | ID | Title | Type | Size | Epic |
|----------|----|-------|------|------|------|
| P0 | PROJ-001 | [title] | feature | S | E01 |
| P1 | PROJ-003 | [title] | bugfix | S | E01 |If no stories have ready status, check for draft stories and suggest running /ideate to refine them, or /backlog-review to assess backlog health.
Define Sprint Goal and Scope
Ask the user to select stories for this sprint and define a sprint goal. Guide them:
- Sprint goal: One sentence describing the outcome (not a list of stories). Example: "Enable users to authenticate via OAuth2"
- Story selection: Based on ready stories, capacity (estimated sessions available), and priority order
- Sizing: S (1 session), M (2-3 sessions), L (3-5 sessions) — based on complexity, not effort hours
- Buffer: Reserve ~15% of sessions for unplanned work
<IF condition="docs/reference/PRD_SUMMARY.md exists">
**Sprint Definition of Done** (derived from PRD): Read PRD Section 6 (NFRs) and Section 7 (scope boundaries). Include applicable thresholds:
- Performance targets from NFRs (e.g., "API < 200ms P95")
- Security requirements from NFRs (e.g., "all PII encrypted")
- Implementation boundaries from Section 7 (Always/Ask first/Never rules)
</IF>Create Sprint Spec
Copy docs/sprints/_TEMPLATE.md to docs/sprints/sprint-<number>.md and fill in:
- Sprint number and goal (from user input above)
- Start date (today)
- Branch name
- Stories table (from selected stories, with sizes)
- Boundaries: Done means, out of scope, risks
- Capacity: available sessions, buffer, constraints
Leave the Decisions, Notes, and Outcome sections empty — they're filled during and after the sprint.
Update progress.md
Update docs/progress.md → ## Current Sprint section with:
- Sprint number and goal
- Branch name and status
- Compact stories table (just #, title, size, status)
- Notes: empty (filled during sprint)
4. Done
Output a summary:
### Sprint Ready
**Sprint [N]: [goal]**
**Branch:** `sprint-<number>`
**Mode:** [Standard / Worktree at ../<path>]
**Main status:** Tests passing, up to date
**Open PRs:** None (or list any that exist)
**Stories:** [count] selected ([total size estimate])
**Sprint spec:** `docs/sprints/sprint-<number>.md`
Ready to start work.
**Next steps:**
1. Clear your context window: `/clear`
2. Start the first story: `/story-cycle [first-story-id]`Always show the actual first story ID from the sprint plan (e.g., /story-cycle E01-S01), not a placeholder.
Skill metrics: Emit a completion event:
echo "{\"type\":\"skill\",\"event\":\"end\",\"skill\":\"sprint-start\",\"outcome\":\"success\",\"ts\":\"$(date -u +%Y-%m-%dT%H:%M:%SZ)\"}" >> docs/sessions/.activity-log.jsonlWhat This Skill Does NOT Do
- Does not load story context (that's
/story-cycle's job) - Does not run analysis agents
- Does not update epic files or backlog
- Does not assume any particular project structure
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

