메뉴
Create a Story Packet for normal/high-risk features. Use after /feature-intake classifies work as normal or high-risk, or when the user asks to break a feature into acceptance criteria, test expectations, and agent-sized work units.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Build a lightweight skill index and load full skill instructions only on demand to reduce startup context pressure.
Select and run a multi-agent workflow pattern for work that exceeds one agent's reliable scope. Use for parallel research, independent reviews, cross-domain changes, or high-risk implementation planning.
Select and run a multi-agent workflow pattern for work that exceeds one agent's reliable scope. Use for parallel research, independent reviews, cross-domain changes, or high-risk implementation planning.
Use composable middleware behaviors for provider/tool execution: retry, caching, timeout, telemetry, and budget guards.
Use this skill whenever the agent makes a mistake, the user observes an avoidable failure, a pattern recurs, or someone says "the agent keeps doing X". Files an "Engineer the Harness" entry — Mitchell Hashimoto's discipline: every failure becomes a permanent prevention mechanism. Always invoke this instead of just fixing the immediate symptom.
Use composable middleware behaviors for provider/tool execution: retry, caching, timeout, telemetry, and budget guards.
| name | create-story |
| description | Create a Story Packet for normal/high-risk features. Use after /feature-intake classifies work as normal or high-risk, or when the user asks to break a feature into acceptance criteria, test expectations, and agent-sized work units. |
| allowed-tools | Read, Write, Edit, Bash(node .claude/skills/create-story/create-story.mjs:*) |
| suggested-turns | 8 |
Turns feature intake output into a concrete .harness/docs/stories/feature-N.md Story Packet.
Run the helper with a title and optional flags:
node .claude/skills/create-story/create-story.mjs "Feature title" --classification=normal --hours=2
If the repo has no obvious test command, pass concrete verification up front:
node .claude/skills/create-story/create-story.mjs "Feature title" --verify-command="npm test -- feature-name"
For high-risk work, declare the implementation layer scope up front:
node .claude/skills/create-story/create-story.mjs "Feature title" --classification=high-risk --hours=4 --layers=service,runtime
Review the generated Story Packet and fill in missing acceptance criteria if needed.
Ensure normal/high-risk features have storyPath, taskContractPath, and
evidencePath in .harness/feature_list.json.
Create or update .harness/task-contracts/<feature-id>.json so the story's
acceptance criteria map to concrete doneRequires gates and an evidence
bundle path.
--verify-command / --regression-command when the default test
command is not specific enough.TBD, TODO, or N/A in task-contract verification fields.Ensure the task contract has a concrete permissions.allow/deny policy.
High-risk work must not use wildcard permissions.
Ensure high-risk task contracts declare scope.allowedLayers. The helper
refuses high-risk stories without --layers; do not create high-risk work
with broad, empty, or placeholder layer scope.
For high-risk work, create or link an ADR and assign a reviewer before implementation.
### Story Packet: <feature-id>
### Path: .harness/docs/stories/<feature-id>.md
### Classification: normal|high-risk
### Task contract: .harness/task-contracts/<feature-id>.json
### Evidence bundle: .harness/evidence/<feature-id>.json
### Next step: approve story | add ADR | implement
scope.allowedLayers for high-risk stories; it drives
task-scoped edit permissions.