一键导入
team-sprint
Run a full development sprint cycle from planning to PR. Use when advancing project with automated skill orchestration.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Run a full development sprint cycle from planning to PR. Use when advancing project with automated skill orchestration.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Root installer anchor for the jaan.to Codex skill pack. Use when installing jaan.to skills in Codex.
Audit AI history and existing workflow, then plan a reliable, evaluated, safe AI system. Use when maturing AI workflows.
Report bugs, feature requests, or skill issues to the jaan-to GitHub repo or save locally. Use when reporting plugin issues.
Detect repeated workflow patterns from AI sessions and suggest skills to automate them. Use when optimizing workflows.
Assemble role-based AI teammates to ship ideas from concept to production via agent teams. Use when orchestrating multi-role delivery.
Orchestrate RED/GREEN/REFACTOR TDD cycle with context-isolated agents. Use when implementing features test-first.
| name | team-sprint |
| description | Run a full development sprint cycle from planning to PR. Use when advancing project with automated skill orchestration. |
| allowed-tools | Read, Glob, Grep, Write($JAAN_OUTPUTS_DIR/**), Bash(git checkout:*), Bash(git pull:*), Bash(git branch:*), Bash(git add:*), Bash(git commit:*), Bash(git push:*), Bash(git log:*), Bash(git diff:*), Bash(gh pr create:*), Bash(gh pr view:*), Bash(cp:*), Skill(pm-sprint-plan), Skill(team-ship), Skill(release-iterate-changelog), Edit(ROADMAP.md), Edit(jaan-to/config/settings.yaml) |
| argument-hint | [cycle-number] [--focus spec|scaffold|code|test|audit] [--tasks 'task1,task2'] |
| disable-model-invocation | true |
| user-invocable | true |
| license | PROPRIETARY |
Full development sprint cycle — plan, execute, verify, and ship via PR.
$JAAN_CONTEXT_DIR/config.md - Project configuration$JAAN_CONTEXT_DIR/boundaries.md - Trust rules$JAAN_TEMPLATES_DIR/jaan-to-team-sprint.template.md - Cycle report template$JAAN_LEARN_DIR/jaan-to-team-sprint.learn.md - Past lessons (loaded in Pre-Execution)$JAAN_CONTEXT_DIR/tech.md - Tech context (if exists)${CLAUDE_PLUGIN_ROOT}/docs/extending/language-protocol.md - Language resolution protocolArguments: $ARGUMENTS
Parse from arguments:
5). If omitted, auto-detect by counting gap-reports/*-cycle/ directories + 1. If no gap-reports exist, default to 1.spec, scaffold, code, test, audit. Passed through to pm-sprint-plan.MANDATORY — Read and execute ALL steps in: ${CLAUDE_PLUGIN_ROOT}/docs/extending/pre-execution-protocol.md
Skill name: team-sprint
Execute: Step 0 (Init Guard) → A (Load Lessons) → B (Resolve Template) → C (Offer Template Seeding)
Also read context files if available:
$JAAN_CONTEXT_DIR/tech.md — Tech stack context$JAAN_CONTEXT_DIR/config.md — Project configurationIf the file does not exist, continue without it.
Read and apply language protocol: ${CLAUDE_PLUGIN_ROOT}/docs/extending/language-protocol.md
Override field for this skill: language_team-sprint
Language exception: Git operations, skill names, YAML, and technical terms remain in English.
ultrathink
Use extended reasoning for:
dev or a cycle branch. If not, ask user to switch.${CLAUDE_PLUGIN_ROOT} resolves and contains skills/. If not → "jaan-to plugin not found. Install or update the plugin."If cycle number provided in arguments, use it.
If not:
# Count existing cycle gap reports
CYCLE_COUNT=$(ls -d gap-reports/*-cycle 2>/dev/null | wc -l | tr -d ' ')
CYCLE_NUMBER=$((CYCLE_COUNT + 1))
If no gap-reports directory exists, default to cycle 1.
CYCLE_BRANCH="cycle/$(printf '%02d' "$CYCLE_NUMBER")"
git checkout dev
git pull origin dev
if git show-ref --verify --quiet "refs/heads/$CYCLE_BRANCH"; then
git checkout "$CYCLE_BRANCH"
else
git checkout -b "$CYCLE_BRANCH"
fi
If the branch already exists (retry after failure), switch to it. Otherwise create it.
Confirm: "On branch $CYCLE_BRANCH (from dev)."
Verify no secrets in tracked files before any commits:
git diff --name-only HEAD 2>/dev/null | tr '\n' '\0' | xargs -0 grep -liE '(sk-|ghp_|token=|password=|api_key=|secret=)' 2>/dev/null || true
If matches found → STOP: "Security risk detected. Remove secrets before proceeding."
Pass through user arguments to pm-sprint-plan:
/pm-sprint-plan [--focus {focus}] [--tasks {tasks}]
Forward --focus and --tasks from the user's original /team-sprint arguments. Omit flags the user did not provide.
This invokes the full pm-sprint-plan workflow:
Wait for pm-sprint-plan to complete. The approved sprint plan artifact will be at $JAAN_OUTPUTS_DIR/pm/sprint-plan/.
After pm-sprint-plan completes, read the sprint plan artifact:
SPRINT_PLAN=$(find "$JAAN_OUTPUTS_DIR/pm/sprint-plan" -name '*.md' -not -name 'README.md' -type f 2>/dev/null | sort -r | head -1)
Verify it contains:
type: sprint-planqueue array with at least 1 itemclosing_skills arrayIf verification fails → "Sprint plan artifact is invalid. Re-run /pm-sprint-plan."
cycle/XX remains. User can re-run /team-sprint to re-plan on the same branch./pm-sprint-plan manually to regenerate, then re-run /team-sprint.Pass the sprint plan to team-ship:
/team-ship --track sprint
team-ship reads the sprint plan artifact and:
Wait for team-ship to complete. This is the autonomous execution phase.
After team-ship completes, stage and commit:
git add -A
git commit -m "$(cat <<'EOF'
feat(cycle): complete cycle {CYCLE_NUMBER} execution
Sprint plan executed via team-sprint.
Skills completed: {completed_count}/{total_count}
EOF
)"
For each successfully completed queue item:
roadmap_ref from the sprint plan- [ ] line in ROADMAP.md- [x] and append commit hash# Example: mark item complete
# - [ ] Implement OAuth login → - [x] Implement OAuth login (abc1234)
Commit ROADMAP changes:
git add ROADMAP.md
git commit -m "docs(roadmap): mark cycle ${CYCLE_NUMBER} items complete"
Create gap report for this cycle at gap-reports/${CYCLE_NUMBER}-cycle/gap-report.md:
# Cycle {CYCLE_NUMBER} Gap Report
## Summary
- Sprint focus: {focus}
- Queue items: {completed}/{total} completed
- Skills executed: {skill_list}
## Completed
{list of completed items with skill and output paths}
## Incomplete
{list of items that failed or were skipped, with reasons}
## New Gaps Discovered
{gaps found during execution that weren't in the original plan}
## Recommendations for Next Cycle
{suggested focus and priority items for cycle N+1}
Commit:
git add gap-reports/
git commit -m "docs(gaps): add cycle ${CYCLE_NUMBER} gap report"
Invoke:
/release-iterate-changelog
Commit changelog updates separately.
git push -u origin "$CYCLE_BRANCH"
gh pr create \
--base dev \
--title "feat(cycle): cycle ${CYCLE_NUMBER} — {focus_summary}" \
--body "$(cat <<'EOF'
## Cycle {CYCLE_NUMBER} Summary
### Sprint Focus
{focus_area} — {bottleneck_stage}
### Execution Queue
{completed_count}/{total_count} items completed
### Skills Used
{skill_list_with_status}
### Changes
{file_change_summary}
### Gap Report
See gap-reports/{CYCLE_NUMBER}-cycle/gap-report.md
---
Sprint plan: $JAAN_OUTPUTS_DIR/pm/sprint-plan/{id}/
Executed via: /team-sprint
EOF
)"
Present PR URL to user.
CYCLE {CYCLE_NUMBER} — COMPLETE
────────────────────────────────
Branch: {CYCLE_BRANCH}
PR: #{pr_number} ({pr_url})
EXECUTION RESULTS
─────────────────
Completed: {completed_count}/{total_count}
Skills: {skill_names}
Focus: {focus_area}
PROGRESS DELTA (if available)
─────────────────────────────
Specification: {before}% → {after}%
Scaffold: {before}% → {after}%
Code: {before}% → {after}%
Tests: {before}% → {after}%
Infra: {before}% → {after}%
NEXT STEPS
──────────
Bottleneck: {next_bottleneck}
Suggested: /team-sprint {CYCLE_NUMBER + 1} --focus {suggested_focus}
"Any feedback on this sprint cycle? [y/n]"
If yes:
/jaan-to:learn-add team-sprint "{feedback}"If no: Cycle complete.