Ship Feature logo

Ship Feature

OrganizationPopular
openonion
ship-feature

Ship a feature end-to-end — update tests, docs, docs-site, then release to PyPI. Use when user says "ship", "ship feature", "release", or asks to publish a new version.

Overview

Publisheropenonion
Repositoryconnectonion
Skill nameship-feature
Stars
1.5K
Forks
218
Bundled files
Instructions only
LicenseApache-2.0
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.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by openonion on GitHub. Read the source before you install it.

Installation

Install the Ship Feature 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.
  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/openonion/connectonion.git /tmp/connectonion
mkdir -p .claude/skills
cp -r /tmp/connectonion/connectonion/useful_skills/ship-feature .claude/skills/ship-feature
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Ship Feature 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 Feature 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 Feature 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 Feature Skill

Ship a feature completely: tests → docs → docs-site → release.

Step 1: Understand What Changed

Read the user's message to identify which feature/module was changed.

Run in parallel:

  • git diff --stat — what files changed
  • git diff — full diff of changes
  • git log --oneline -5 — recent commit context

Step 2: Update Tests

Find the relevant test file:

  • glob("tests/**/*.py") — find all test files
  • Match test file to changed source file (e.g. src/agent.pytests/unit/test_agent.py)

Update the test file:

  • Add or update test cases that cover the new behavior
  • Run tests to confirm they pass: python -m pytest tests/unit/test_<module>.py -v
  • If tests fail, fix them before proceeding

Step 3: Update docs/

This step is required. Do not skip.

Find ALL docs that need updating:

bash
glob("docs/**/*.md")

For each changed area:

  • If a doc file exists for it — update it with the new behavior, params, examples
  • If no doc file exists — create one (look at neighboring files for format)
  • Also check index/README files (e.g. docs/cli/README.md, docs/useful_tools/README.md) — update the table of contents if you added something new

Commit docs/ changes as part of the release commit (not separately).

Step 4: Update docs-site

This step is required. Do not skip even if docs-site/ is not present locally.

docs-site is a separate Next.js git repo. Check if it's cloned:

bash
ls docs-site/

If docs-site/ exists:

  • Find the corresponding page: glob("docs-site/app/**/*.{tsx,mdx}")
  • Update it to match what you changed in docs/
  • Respect existing component structure (CommandBlock, CodeBlock, etc.)
  • Run its lint and production build. Prepare the docs commit, but do not publish version availability before the matching PyPI package and GitHub Release are public. Publish the prepared docs in Step 5e.

If docs-site/ does NOT exist locally:

  • Tell the user explicitly: "docs-site was not updated — clone it and run co copy ship-feature --force to re-run"
  • Do NOT silently skip — the user must know this is incomplete

Step 4b: Record the design and release story

For a feature-train launch, first beta, first RC, stable release, or material architecture/workflow decision, create or substantially update a Design Journal post in the docs site. Maintenance-only patches need release notes unless they contain a reusable lesson.

The post must explain:

  • the problem and user impact;
  • alternatives considered;
  • the decision and tradeoffs;
  • evidence and current limitations;
  • what would make the team revisit the decision.

Keep one canonical Markdown source and a rendered blog page. Add the post to the blog index, internal navigation, site search, dynamic and static sitemaps, llms.txt, and relevant AI-readable indexes. Give it unique metadata, a canonical URL, social metadata, and TechArticle or BlogPosting structured data. Test desktop and mobile layouts. Draft with provisional wording; do not claim that an artifact is published until that is verified.

Step 5: Release

5a. Determine new version

Find and read the current version — check these locations in order:

bash
grep -r "__version__" --include="*.py" -l   # find which file has version
cat pyproject.toml | grep "^version"         # or pyproject.toml
cat setup.py | grep "version="               # or setup.py

Apply versioning rules (read VERSIONING.md if it exists, otherwise use semver):

  • Default to PATCH +1 for normal shipped work, including small user-facing improvements
  • Use MINOR only when the user explicitly asks for it or the change is clearly a larger compatibility-safe feature release
  • Use MAJOR only for breaking changes, stable-release milestones explicitly requested by the user, or required rollover rules
  • If VERSIONING.md exists, follow its rollover rules exactly, but do not jump to a larger bump unless the rules require it

5b. Update version in all files that contain it

Search for every file containing the current version string and update each one:

bash
grep -r "X.Y.Z" --include="*.py" --include="*.toml" --include="*.cfg" -l

Common locations: __init__.py, pyproject.toml, setup.py, setup.cfg

5c. Validate and prepare the release commit

Remove old build output, build the candidate once locally, and validate only the two exact versioned artifacts. This is validation, not publication.

bash
rm -rf dist/
python -m build
python -m twine check dist/<package>-X.Y.Z.tar.gz dist/<package>-X.Y.Z-py3-none-any.whl

Then stage only what changed — do NOT blindly stage all files:

bash
git add -p   # or stage specific files that were actually modified
git status   # confirm what's staged before committing
git commit -m "Release vX.Y.Z: <feature description>"
git push -u origin <release-branch>
gh pr create --title "Release vX.Y.Z: <feature description>"

Do not tag an unreviewed branch. After the release PR is reviewed and merged, fetch the target branch, resolve the exact merge commit, and create one annotated, immutable vX.Y.Z tag that points to that commit.

bash
git fetch origin
git tag -a vX.Y.Z <reviewed-merge-commit> -m "Release vX.Y.Z"
git push origin vX.Y.Z

5d. Let the reviewed tag workflow publish

Pushing the tag starts .github/workflows/release.yml. Wait for that exact run to rerun the matrix, build once, publish through PyPI Trusted Publishing, verify the public artifacts, and create the GitHub Release. A manual dispatch may retry the existing tag; it must never select an arbitrary branch. Never publish package bytes from the workstation or race the workflow with a second registry writer.

bash
gh run list --workflow release.yml --limit 1
gh run watch <run-id> --exit-status

Confirm that the exact PyPI version and GitHub Release are public and that a preview is marked Prerelease rather than Latest before publishing documentation.

5e. Publish documentation and the Design Journal

After the exact PyPI package and GitHub Release are public, commit and push only the reviewed docs-site files. Verify the deployed stable/preview labels, installation commands, canonical blog URL, structured data, sitemap entry, internal links, AI-readable indexes, and mobile rendering. If the docs site cannot be published, report the release handoff as incomplete rather than silently skipping it.

Checklist

  • Tests updated and passing
  • docs/ updated
  • docs-site/ and any required Design Journal post updated; lint and build pass
  • Version bumped in every file that held it, and they agree (in connectonion: connectonion/_version.py and pyproject.toml; __init__.py only re-exports it and there is no setup.py)
  • Release PR reviewed and merged
  • Immutable tag points to the reviewed merge commit
  • Exact-tag release.yml run passed
  • Exact PyPI package and GitHub Release verified public
  • Docs-site version state and Design Journal published after public artifacts were verified

Notes

  • docs/ and docs-site are both required — never silently skip either
  • If docs-site is missing locally, warn the user instead of skipping
  • If the user says "skip release", stop after docs-site
  • If the user specifies a version explicitly, use that instead of auto-calculating
  • Never force-push or amend published commits
  • Never publish package artifacts directly from a workstation

Frequently asked questions

What does the Ship Feature AI skill do?

Ship a feature end-to-end — update tests, docs, docs-site, then release to PyPI. Use when user says "ship", "ship feature", "release", or asks to publish a new version.

Why use Ship Feature on TypingMind?

Because you install it once and use it with any model. Ship Feature 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 Feature in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/openonion/connectonion/tree/main/connectonion/useful_skills/ship-feature. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Ship Feature?

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 Feature?

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

Is the Ship Feature AI skill free?

Yes. It is published on GitHub by openonion under the Apache-2.0 license. 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 👇