MCPInstall

ship-it

Use when asked to ship a completed change, commit and open a PR, or push changes and watch CI. Use babysit for monitoring an existing PR.

Installs
1.7K
Last 24h
—
Trending rank
—
Security
Not audited

Install

Works with every agent the skills CLI supports.

npx skills add https://github.com/jssblck/agents --skill ship-it
  1. 01

    Install

    Run the command in your project. The CLI asks which agents to install for.

  2. 02

    Describe the task

    Ask your agent for something this skill covers. It loads ship-it automatically when relevant.

  3. 03

    Review the result

    The agent follows the SKILL.md steps, scripts and references bundled with the skill.

SKILL.md

The instructions your agent reads when the skill is triggered.

Ship it

For a full shipping request, finish with the change committed, pushed, open as a ready pull request, and green in CI. Open a new PR as non-draft, then finish the repository's required review and local verification while CI runs.

A narrower request keeps its scope. For CI-only repair, fix the affected branch and checks without creating a PR or changing its draft/ready state, base, title, or unrelated content. For monitoring only, report results without making fixes. An explicit request to keep a draft takes precedence over the full workflow.

Do not merge the PR unless asked. Pause only for material ambiguity, a required approval, or a blocker that cannot be resolved with available access.

Commit the requested work

Work on a feature branch, not the default branch. Stay on the current feature branch when it owns the change; otherwise follow the repo's naming convention, or use agent/ when none exists.

Inspect the diff and working tree. Preserve unrelated user changes and stage explicit paths. Group related files into logical commits.

Use an imperative subject. Explain the reason for the change where it is not obvious from the subject. Follow active attribution and punctuation instructions. Wrap commit bodies around 80 columns, but do not hard-wrap PR paragraphs.

Open or reuse the PR

For a full shipping request, push and check for an existing PR:

gh pr list --state open --head <branch> --json number,title,url,isDraft,baseRefName,headRefName

Reuse the existing PR. If it is a draft, mark it ready with gh pr ready unless the user asked to keep it a draft. If no PR exists, create a non-draft PR with gh pr create. Use --draft only when the user explicitly requests a draft. Use the default branch as the base unless this is a dependent layer in an existing stack or the user specified another base.

Explain why the change is needed and what changed. Reference the originating issue with Fixes #<n> when applicable. Preserve existing context when editing the body. Do not add labels or reviewers unless requested.

Run configured review and verify proportionally

For a full shipping request, run the project's signoff workflow if it defines one, after opening or reusing the PR and setting its intended readiness. Use the repository's current review policy; do not invent a gate when none exists.

Read the repo's check commands. Run required local checks and the tests that cover the changed behavior. Generate required code before building.

  • Include integration tests when the change reaches that boundary. Confirm relevant tests ran rather than silently skipped.
  • Existing automated end-to-end tests can supply runtime evidence. Do not repeat their coverage manually without a distinct risk to check.
  • Use manual interaction for uncovered integration or visual risks. Exercise representative affected flows rather than every route sharing a component. Check responsive layouts when the change affects them.
  • For CLI or API behavior, a focused test or real command/request can supply evidence. Choose the method that catches the risk.
  • For changes with no runtime surface, state that no runtime capture was needed.
  • Report local environment gaps and let CI cover checks that require its platform.

After passing checks, repeat or broaden verification only for new edits, failures, or unresolved risks. Do not refactor unrelated code to improve the verification process. Commit and push any necessary fixes.

Add evidence

For a full shipping request, add a Testing section with exact commands, results, and relevant gaps. Use existing automation output when sufficient.

Inline concise text evidence in a language-tagged fence: console for a command transcript, json for JSON, or text for other output. Remove secrets and private data from proof before publishing it.

Attach useful screenshots or recordings with gh --attach, not gh-image. Follow github-image-upload for supported media, attachment syntax, and verification. No media is required when text evidence covers the change. Do not commit proof files.

Invoking this full workflow authorizes adding its proof to the PR. Keep evidence in the PR body, preserve its existing explanation, and keep any required attribution footer last.

Verify the intended PR base, head, state, and evidence with gh pr view.

Babysit the published PR

After publishing and verifying the PR, use babysit to check for merge conflicts before waiting and watch CI through completion. Watch the complete check set so a filtered view cannot hide a failure. For monitoring-only requests, report conflicts without resolving them.

For failures within the request, diagnose, fix, run affected checks, commit, push, and watch again. Refresh proof when the behavior it documents changes. Preserve required checks; do not bypass them or claim success when checks failed, were cancelled, or did not run as expected.

Report unrelated failures rather than expanding the work without direction. If access or infrastructure blocks progress, identify the blocker and checks that remain unresolved.

Report

Link the PR and state its actual readiness and CI results. Summarize meaningful changes and verification gaps. Say CI is green only when the complete applicable check set supports that claim.

FAQ

What does the ship-it skill do?

Use when asked to ship a completed change, commit and open a PR, or push changes and watch CI. Use babysit for monitoring an existing PR.

How do I install ship-it?

Run npx skills add https://github.com/jssblck/agents --skill ship-it in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/ship-it/ (personal) or .claude/skills/ship-it/ (project).

Which AI agents can use ship-it?

Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.

Is ship-it safe to install?

No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.

Who maintains ship-it?

It is published by jssblck in jssblck/agents. It has 1,663 recorded installs.

More from jssblck/agents

View all →

using-sops

jssblck/agents

Use for secrets or key setup in a repo with .sops.yaml, encrypted environment files, or a pnpm secrets script.

Security1.7K installs

code-craft

jssblck/agents

Apply Jess's engineering conventions to invariants, boundary typing, state, effects, proofs, error handling, and abstraction decisions.

1.7K installs

merge-open-prs

jssblck/agents

Use when asked to merge all open PRs or land a specified set of PRs together.

1.7K installs

tag-release

jssblck/agents

Use when asked to tag or cut a patch, minor, or major release.

1.7K installs

gh-stack

jssblck/agents

Use when working in a checked-out PR stack, handling dependent PRs or branch layers, or planning to split work into stacked PRs.

parallel-agents

jssblck/agents

Use when agents work concurrently in a codebase, shared files have recurring merge conflicts, or AGENTS.md or CLAUDE.md needs maintenance.

Popular code quality & workflow skills

View category →

grill-me

mattpocock/skills

A relentless interview to sharpen a plan or design.

Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.

handoff

mattpocock/skills

Compact the current conversation into a handoff document for another agent to pick up.

domain-modeling

mattpocock/skills

Build and sharpen a project's domain model. Use when discussing codebase terminology, writing or editing a GLOSSARY.md, or recording or editing an ADR.

codebase-design

mattpocock/skills

Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.

diagnosing-bugs

mattpocock/skills

Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.