Ship logo

Ship

Organization
nightly-labs
ship

Ship the current branch to main. Review the diff, fix findings, open the pull request, babysit the NorbiAI review and CI until green, then squash-merge. Use when a feature is done and the user asks to ship it, open and merge a PR, or get the change to main.

Overview

Publishernightly-labs
Repositoryopenbot
Skill nameship
Stars
521
Forks
81
Bundled files
1
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 1 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by nightly-labs on GitHub. Read the source before you install it.

Installation

Install the Ship AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/nightly-labs/openbot/tree/main/.agents/skills/ship
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/nightly-labs/openbot.git /tmp/openbot
mkdir -p .claude/skills
cp -r /tmp/openbot/.agents/skills/ship .claude/skills/ship
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Ship in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Ship on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Ship is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Ship

Move one finished branch to main through five gates: local review, pull request, NorbiAI review, CI, merge. Do each gate in sequence. Do not skip a gate to go faster.

When the user invokes this skill, they approve the merge at the end, but only when every gate is green. Stop and ask when a gate needs a product decision, a force-push, or a bypass.

Before you start

  • Read AGENTS.md and each nested AGENTS.md for a directory in the diff.
  • Run git status and git branch --show-current. When the branch is main, create a branch first. Do not ship from main.
  • Run git fetch origin main. Get the change set with git diff --stat origin/main...HEAD and git status --short.
  • When a pull request for the branch exists (gh pr view --json number,state,url), continue from the first gate that is not done.
  • Keep a checklist in TASKS.md with one item for each gate.

Gate 1 — Review the diff locally

A local review is less expensive than a NorbiAI round. Find the problems before the push.

  1. Commit the finished work. Write the message in ASD-STE100 Simplified Technical English. Leave unrelated changes out of the commit.
  2. Review origin/main...HEAD for correctness, data loss, the non-negotiable rules in AGENTS.md, and reuse. Use the code-review skill when it is available, or a review subagent with the diff range.
  3. Triage each finding as babysit Step 2 says: fix only findings that you can prove. Record the others as not valid, with the reason.
  4. Run the narrowest relevant test file and biome check --write --max-diagnostics=none <changed paths>. Run one typecheck project for the code you changed, as AGENTS.md says. Do not run broad checks.
  5. Commit the fixes. The pre-commit hook runs its own checks. Do not use --no-verify. When the hook fails, fix the cause and commit again.

For a UI change, run the smoke skill and take before and after screenshots. Upload them with gh pr create --attach or gh pr edit --attach (gh 2.101 or later; run gh --version). Do not commit them.

Gate 2 — Open the pull request

  1. Push the branch: git push -u origin HEAD.
  2. Write the body from .github/PULL_REQUEST_TEMPLATE.md. Fill ## What changed with the user-visible result and the surfaces the change touches. Fill ## Verification with the checks you ran and the checks you left to CI. Remove the sections the template says to remove.
  3. State the model and harness in the body.
  4. Create the PR: gh pr create --base main --title "<title>" --body-file <file>. Write the title as one imperative sentence, as recent main commits do.
  5. When the link_pull_request tool is available, link the PR URL.

Gate 3 — Babysit the NorbiAI review

Run the babysit skill on the PR. It owns the review loop: triage, fixes, rebuttals, and re-review requests. Continue to Gate 4 only when the NorbiAI review status on the head SHA is success.

Gate 4 — Make CI green

The branch protection on main requires the Check status and resolved conversations. It does not require the branch to be up to date with main: a PR that is behind main can merge when it has no merge conflict.

  1. Watch the checks on the head SHA: gh pr checks <pr> --watch --interval 120. Run it in the background.
  2. For each failed job, read only the failed log: gh run view <run-id> --log-failed. Find the cause.
  3. When the failure comes from the change, fix it, commit, and push. Each push starts a new NorbiAI review. Go back to Gate 3 for that review.
  4. When the failure is not related to the change, such as a network timeout, run gh run rerun <run-id> --failed once. When the same job fails again, stop and report it to the user.
  5. Do not update the branch only because main moved. When gh pr view <pr> --json mergeable shows CONFLICTING, merge main into the branch, resolve the conflicts, and push. Do not force-push. The merge starts a full NorbiAI review. Go back to Gate 3.
  6. Answer each open human review thread. Do not resolve a thread that another person opened before you answer it.

Never weaken or skip a test to make CI green.

Gate 5 — Merge

Merge only when all of these are true on the same head SHA:

  • gh pr view <pr> --json mergeStateStatus,reviewDecision,statusCheckRollup shows mergeStateStatus CLEAN. DIRTY means a merge conflict: go back to Gate 4, step 5.
  • Check and NorbiAI review are success.
  • No P0 or P1 finding remains, and no conversation is open.

Then merge:

gh pr merge <pr> --squash --match-head-commit <head-sha>

--match-head-commit stops the merge when a new push arrived after your last check. The repository deletes the remote branch after the merge.

Do not use --admin. Do not bypass branch protection. Do not delete this worktree or the local branch.

After the merge:

  1. Run gh pr view <pr> --json state,mergeCommit and confirm MERGED.
  2. Run git fetch origin main and confirm that origin/main contains the merge commit.

Stop and ask

Stop and ask the user when:

  • A fix needs a force-push, a history rewrite, or a change outside this repository.
  • A valid finding needs a product decision.
  • The babysit skill stops with no progress.
  • CI fails twice on a cause that you cannot fix in the change.
  • The merge needs --admin or a protection change.

Report format

Keep the report short:

GateResultEvidence
Local reviewfindings fixed / rebuttedcommit SHAs
Pull requestopenedPR URL
NorbiAIsuccess after N roundsreview comment link
CIsuccessrun link
Mergemergedmerge commit SHA on main

Then list the checks you ran locally and the checks you left to CI, one line each.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Ship AI skill do?

Ship the current branch to main. Review the diff, fix findings, open the pull request, babysit the NorbiAI review and CI until green, then squash-merge. Use when a feature is done and the user asks to ship it, open and merge a PR, or get the change to main.

Why use Ship on TypingMind?

Because you install it once and use it with any model. Ship is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Ship in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/nightly-labs/openbot/tree/main/.agents/skills/ship. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Ship?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Ship?

As many as you like. As long as a model supports skills, you can use Ship with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Ship AI skill free?

It is published on GitHub by nightly-labs. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇