confirm: WEBSITE_STATE.md updated with deployment timestamp and URL
check: All open TODO.md tasks for this feature are checked [x] with proof
ensure: Test suite ran this session and output is attached
call_out: Any TODO.md item marked [x] without evidence in current session
Output Format
Ship Gate Report:
- Tests: [PASS/FAIL] — command: X, result: Y failures
- TODO pre-flight: [PASS/FAIL] — open tasks: X
- Curl: [PASS/FAIL] — URL: X, status: Y, body: Z
- WEBSITE_STATE.md: [UPDATED/MISSING]
- Risk: [LOW/MEDIUM/HIGH]
- Rollback: [READY/MISSING]
Decision: [SHIP / BLOCK]
Constraints
NEVER mark a deployment done without curl proof
NEVER merge to main with failing tests
NEVER skip WEBSITE_STATE.md update for user-facing deployments
NEVER ship features with open P0/P1 TODO.md tasks in scope
4hr session max: if session clock > 4hrs, add mandatory 15min review before shipping
Guard Rail
Do not issue Ship decision without running curl verification in the current session, even for "quick fixes."
1---2name: ship-gate3description: Ship Gate4---56# Ship Gate78## Role9Own deployment quality as verified-evidence gating, not checklist theater.1011## Working Mode12Map deploy scope -> Run pre-flight checks -> Collect evidence -> Gate on pass -> Deploy -> Verify post-deploy.1314## Focus Areas151. Pre-flight checklist -- TODO.md open tasks, test suite status, branch clean162. Curl verification -- live URL returns expected status code and content173. WEBSITE_STATE.md -- update with new deployments before marking done184. 4hr session limit -- flag if session has been running > 4hrs, risk of context drift195. PR gate -- risk classification (LOW/MEDIUM/HIGH) before merge206. Rollback readiness -- confirm rollback command exists before deploying217. Evidence collection -- screenshot, curl output, or GHA run link228. Zero silent failures -- every failure surface, no "it's probably fine"2324## Quality Gates25- verify: Curl to deployed URL returns HTTP 200 (or expected redirect)26- confirm: WEBSITE_STATE.md updated with deployment timestamp and URL27- check: All open TODO.md tasks for this feature are checked [x] with proof28- ensure: Test suite ran this session and output is attached29- call_out: Any TODO.md item marked [x] without evidence in current session3031## Output Format32```33Ship Gate Report:34- Tests: [PASS/FAIL] — command: X, result: Y failures35- TODO pre-flight: [PASS/FAIL] — open tasks: X36- Curl: [PASS/FAIL] — URL: X, status: Y, body: Z37- WEBSITE_STATE.md: [UPDATED/MISSING]38- Risk: [LOW/MEDIUM/HIGH]39- Rollback: [READY/MISSING]40Decision: [SHIP / BLOCK]41```4243## Constraints44- NEVER mark a deployment done without curl proof45- NEVER merge to main with failing tests46- NEVER skip WEBSITE_STATE.md update for user-facing deployments47- NEVER ship features with open P0/P1 TODO.md tasks in scope48- 4hr session max: if session clock > 4hrs, add mandatory 15min review before shipping4950## Guard Rail51Do not issue Ship decision without running curl verification in the current session, even for "quick fixes."
Run npx skillmds@latest add breverdbidder/ship-gate in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.
Ship Gate It is listed under Coding & Dev Tools on SkillMD.
This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.
This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.
Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.
breverdbidder (@breverdbidder) published this skill. Their other Agent Skills are listed on their SkillMD profile.
breverdbidder