ship
Release the current state of the project — pre-flight checks, version bump and tag, changelog entry, deploy, and an optional announcement draft.
Skill by RubenGlez
Chalk Labs
Need help deploying ship?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/rubenglez-ship
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: testing, devops
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/ship
What's inside
- Skill name
ship- Description
- Release the current state of the project — pre-flight checks, version bump and tag, changelog entry, deploy, and an optional announcement draft. Use after qa has passed and docs are updated, or when the user wants to ship, release, deploy, publish, or cut a version.
- License
- MIT
- Instructions
- 74 lines in SKILL.md · view on GitHub
Before installing
- Instructions call the network
Files (2)
- REFERENCE.md
- SKILL.md
SKILL.md (first lines)
# Ship Shipping is outward-facing and hard to reverse. Never deploy, push a tag, or publish without explicit user approval of the release plan (Step 3). ## Step 1: Pre-flight Verify the project is releasable: - `.harness/qa/report.md` — the latest report should have no open issues. If it does, list them and ask whether to ship anyway. - `git status` — working tree must be clean and on the main (or designated release) branch. Uncommitted changes block the release. - If `.doctier.yml` exists: run `doctier check`. A non-zero exit means a misclassified or plaintext-committed doc — this blocks the release. - Run the project's static checks and test suite one final time. Any failure blocks the release. - For **tag-triggered pipelines** (EAS Workflows, GitHub Actions releases, etc.): confirm the most recent pipeline run on the main branch passed before tagging. A broken pipeline consumes a version number — every retry needs a new tag. If the last run failed, investigate and fix the pipeline config in a commit *before* the release commit, then verify it passes. Where the tooling supports it, validate the pipeline config locally before committing (e.g. `eas build --dry-run` for EAS, schema linters for GitHub Actions) — a config validation error caught locally is cheaper than a burned version number. - **License**: a public repo should carry a license file. If no `LICENSE` / `LICENSE.md` / `LICENSE.txt` exists, offer to add one — recommend MIT (fill in the author name and current year) but let the user choose another or decline. Never pick a license silently; it is a legal decision. This does not block the release — if the user declines, proceed and record the choice in `AGENTS.md` under "Release" (e.g. "intentionally unlicensed") so future runs don't re-ask. If any check fails, stop and report — do not work around it. ## Step 2: Determine the release procedure Look for an existing procedure, in order: 1. `AGENTS.md` — a "Release" or "Deploy" section from a previous /ship run 2. Deploy config in the repo: `vercel.json`, `fly.toml`, `netlify.toml`, `Dockerfile` + compose, `eas.json` / fastlane, `publishConfig` in package.json, a `deploy` script or Makefile target, a deploy workflow in `.github/workflows/` 3. If nothing is found: ask the user what "deploying" means for this project (command, platform, manual steps), then record the answer in `AGENTS.md` under a "Release" section so future runs skip this question. ## Step 3: Propose the release plan Determine the version bump from what shipped since the last tag (`git log <last-tag>..HEAD` and the feature specs): breaking → major, features → minor, fixes only → patch. If there is no version scheme yet, propose starting at `0.1.0`. Present the plan and wait for approval: > Releasing **v1.3.0**: > - Features: [list from specs/commits] > - Fixes: [list] > - Deploy: [procedure from Step 2] > - Changelog: [will update CHANGELOG.md / will include notes in tag message] > > Proceed? ## Step 4: Version, changelog, tag After approval: 1. Bump the version in the manifest (`package.json`, `pyproject.toml`, `Cargo.toml`, ...). 2. If `CHANGELOG.md` exists, prepend the entry (see [REFERENCE.md](REFERENCE.md) for format). If it doesn't exist, put the release notes in the tag message instead — don't create new doc files unasked. 3. Commit the bump, create an annotated tag `v[X.Y.Z]`, and push branch + tag. ## Step 5: Deploy and verify Run the procedure from Step 2. Then verify the release where it actually runs — pick the strongest available signal: - Web: hit the production URL or health endpoint; load the key page that shipped - API: curl the live endpoint for one shipped behavior - Package: install the published version in a temp dir and import/run it - Mobile: confirm the build was submitted/accepted by the store tooling A deploy command exiting 0 is not verification. If verification fails, report immediately and ask whether to roll back — do not retry blindly. ## Step 6: Record the release - Update `.harness/product/roadmap.md`: mark the shipped phase/features with the version and date. - If the deploy procedure differed from what `AGENTS.md` documented, update that section. ## Step 7: Announcement draft (optional) Offer once: a short announcement based on the changelog entry — user-facing benefits, not implementation details (see [REFERENCE.md](REFERENCE.md) for the template). Skip silently if the user declines. ## Step 8: Recommend next step - Shipped clean: "Run /implement for the next phase, or /ideate if the roadmap is done." - Shipped with known open issues: list them as candidates for the next phase. - Verification failed: the rollback/fix decision comes first; recommend /qa after.
From RubenGlez/harness, MIT.
More from RubenGlez/harness
- 0.552score
metrics
RubenGlez
Measure traction and discoverability across all non-archived GitHub repos and their publications (npm, App Store, Play Store, web properties), judge each…
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
task
RubenGlez
Make a small change to a shipped product — a bug fix, behavior tweak, or micro-feature — as one verified slice, syncing the affected feature spec on the way…
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
migrate-docs
RubenGlez
Discover all documentation in the repo, classify each file, transform content to match harness templates, and migrate everything to the correct location.
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
next-step
RubenGlez
Recommends the next harness skill to use by inspecting the repo, docs, git state, and recent work.
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
prototype
RubenGlez
Build throwaway code to answer a specific design question before committing to an implementation approach.
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
write-a-skill
RubenGlez
Create new harness skills with proper structure, progressive disclosure, and bundled resources.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
product-plan
RubenGlez
Define the full product vision — audience, positioning, features, roadmap, and UX direction — through a structured interview.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
qa
RubenGlez
Test implemented features against their acceptance criteria, fix simple failures, and write a QA report.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
dev-plan
RubenGlez
Analyze product docs and the codebase to decide architecture, tech stack, and implementation approach, then generate a technical spec for every must-have…
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
update-docs
RubenGlez
Update all project documentation to reflect the current state of the codebase.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
implement
RubenGlez
Classify planned features as autonomous or human-approval-required, then implement the current phase as parallel vertical slices — one subagent per feature.
SkillBFreshness: B★ 1Claude Code +9 - 0.497score
address-cves
RubenGlez
Find and fix high and critical application dependency CVEs across all non-archived GitHub repositories available to the authenticated gh CLI user.
SkillCFreshness: C★ 1Claude Code +9
Signals
- Stars
- 1
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Jul 1, 2026 (4 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/RubenGlez/harness
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from RubenGlez
- 0.552score
metrics
RubenGlez
Measure traction and discoverability across all non-archived GitHub repos and their publications (npm, App Store, Play Store, web properties), judge each…
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
task
RubenGlez
Make a small change to a shipped product — a bug fix, behavior tweak, or micro-feature — as one verified slice, syncing the affected feature spec on the way…
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
next-step
RubenGlez
Recommends the next harness skill to use by inspecting the repo, docs, git state, and recent work.
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
migrate-docs
RubenGlez
Discover all documentation in the repo, classify each file, transform content to match harness templates, and migrate everything to the correct location.
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
prototype
RubenGlez
Build throwaway code to answer a specific design question before committing to an implementation approach.
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
write-a-skill
RubenGlez
Create new harness skills with proper structure, progressive disclosure, and bundled resources.
SkillBFreshness: B★ 1Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get ship running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
ship skill. Agent Mart. https://agentmart.dev/skills/rubenglez-ship (last updated 2026-10-24)