Package Release logo

Package Release

Community
dannote
package-release

Prepare, publish, repair, or verify package releases and changelogs. Always use for package/version release, publish, ship, tag, changelog, release notes, registry publication, documentation publication, or GitHub Release work. Distinguish package releases from application deployments before acting.

Overview

Publisherdannote
Repositorydot-pi
Skill namepackage-release
Stars
57
Forks
4
Bundled files
Instructions only
LicenseMIT
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 dannote on GitHub. Read the source before you install it.

Installation

Install the Package Release 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/dannote/dot-pi/tree/master/skills/package-release
  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/dannote/dot-pi.git /tmp/dot-pi
mkdir -p .claude/skills
cp -r /tmp/dot-pi/skills/package-release .claude/skills/package-release
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Package Release 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 Package Release 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 Package Release 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.

Package Releases

Release notes are user documentation. CHANGELOG.md is their canonical source. Never invent a second narrative during publication.

Classify the request

  • Update the changelog — edit Unreleased only. Do not bump, tag, push, or publish.
  • Prepare a release — bump/align versions, roll the changelog, validate, and stop before irreversible actions.
  • Release / publish / ship — complete every applicable release output: package, package documentation, tag, GitHub Release, and public verification. Do not stop after publishing only one artifact.
  • Repair a release — derive missing or corrected metadata from the tagged changelog; do not generate replacement prose.
  • Deploy — this skill does not define application deployment. Clarify when “release” could mean an OTP/app deployment rather than a package version.

A direct request to release, publish, ship, or cut a version authorizes the complete workflow after a clean preflight. Ask only when the version or intent is ambiguous, repository state is unexpected, project instructions require confirmation, or authentication/2FA is needed.

Non-negotiable release-note contract

  • GitHub Release title is exactly the tag, normally vX.Y.Z.
  • GitHub Release body is exactly the body of the matching CHANGELOG.md version section, excluding the level-two version heading.
  • Do not add an opening summary, closing sentence, generated prose, commit list, contributor list, checksum, or publication status.
  • Do not add Install/Installation sections or dependency snippets.
  • Do not append Hex, HexDocs, npm, crates.io, changelog, compare, marketing, or “full changelog” links.
  • Never use generated GitHub notes (--generate-notes).
  • Never embellish terse changelog notes.
  • Always pass multiline notes through a file with --notes-file; never inline Markdown in --notes or a shell string.
  • Never use regex, awk, or sed to determine Markdown section boundaries. Read the Markdown structure directly or use the repository’s canonical parser/extractor.
  • Preserve directly relevant issue, pull-request, migration, or security links already present in the canonical changelog. Do not add links during release publication.

Changelog audience

Every entry must answer:

What changed for a package user deciding whether or how to upgrade?

Include:

  • New public behavior or APIs
  • Changed behavior
  • User-visible bug fixes
  • Observable performance improvements
  • Compatibility and platform changes
  • Deprecations, removals, security changes, and migration requirements

Exclude:

  • Test additions, test counts, CI status, or quality-gate output
  • Corpus, calibration, benchmark-gate, precision-ratchet, or release-gate results
  • Checksums, publication status, and generated documentation status
  • Local paths, path dependencies, dogfooding setup, and maintainer workstation state
  • Release preparation and internal workflow changes
  • Internal refactors without a user-visible outcome
  • Dependency/lockfile updates without compatibility, security, or behavioral impact
  • Temporary operational details and implementation-specific validation

Describe internal work through its user-visible outcome. For example, write “Improved analysis performance on large diffs,” not how many packages or tests were used to validate it.

Canonical changelog format

Use this format for new owned-project release sections unless repository instructions or an upstream fork require preserving another established format:

markdown
# Changelog

## Unreleased

## 1.2.3 - 2026-07-20

### Breaking changes

- ...

### Added

- ...

### Changed

- ...

### Deprecated

- ...

### Removed

- ...

### Fixed

- ...

### Security

- ...

### Compatibility

- ...

Rules:

  • Version headings do not include v; tags normally do.
  • Dates use ISO YYYY-MM-DD.
  • Keep releases in reverse chronological order.
  • Keep an Unreleased section at the top.
  • Omit empty categories.
  • Use only the categories above, in the order above.
  • Put upgrade-required source/config/data changes under Breaking changes, including for 0.x packages. Do not duplicate them under Changed or Removed.
  • Use Compatibility only for supported Elixir/OTP/runtime/framework versions, operating systems, architectures, or explicit interoperability boundaries—not ordinary dependency bumps.
  • Do not use New, Improved, Highlights, Fixes, What's New, Tests, Validation, Links, or project-specific release headings.
  • Do not rewrite historical sections merely to adopt the standard. Normalize Unreleased and future releases prospectively.
  • Preserve upstream changelog conventions in maintained forks, while still enforcing the audience and exact-release-body rules.

Discover project release mechanics

Before choosing commands, inspect:

  1. Repository instructions (AGENTS.md, contributing/release docs)
  2. Working-tree and remote synchronization state
  3. Current package version(s) and intended tag
  4. CHANGELOG.md and the exact target section
  5. Package manifests and package file lists
  6. Tag-triggered workflows and reusable release workflows
  7. Existing tag and GitHub Release state
  8. Registry authentication and required 2FA

Respect project-specific publication mechanics. They may change command order; they do not override the release-note audience contract.

Common topologies:

  • Manual registry release — validate, publish package/docs, push the correct tag, create the GitHub Release.
  • Tag-triggered publication — push the prepared commit and tag, watch the workflow, then verify every output. Do not duplicate local publication.
  • Precompiled/native artifacts — tag may create the GitHub asset release first. Wait for required assets/checksums, complete package publication, then edit the existing GitHub Release to the canonical title/body instead of creating another release.
  • Monorepo — determine version alignment, package-specific versus shared tags/changelogs, and every package expected to publish before acting.

Never force-move a release tag unless the user explicitly approves and the project’s release process permits it. Prefer a corrective patch release when immutable artifacts may already exist.

Prepare and review notes

Before the first irreversible action:

  1. Read the target level-two changelog section.
  2. Copy only its body to /tmp/<package>-<version>-release-notes.md.
  3. Read the temporary file back.
  4. Reject it if it violates the audience or heading rules.
  5. Confirm the package version, commit, tag, release topology, exact title, and exact body.

Show the exact body when it was drafted or changed during the current task. Do not present a separate model-written summary as proposed release notes.

Publish

Use repository-specific commands. For GitHub, use the notes file:

bash
gh release create "$tag" \
  --verify-tag \
  --title "$tag" \
  --notes-file "$notes_file"

If an artifact workflow already created the release:

bash
gh release edit "$tag" \
  --title "$tag" \
  --notes-file "$notes_file"

Set prerelease/latest state according to the version and project policy. Do not add body text to communicate those states.

If a registry requests 2FA, stop and ask for the current code. Do not retry stale codes or claim publication succeeded after an authentication failure.

Verify before declaring success

A release is complete only after verifying every applicable item:

  • Registry version exists and has the intended metadata
  • Package documentation exists
  • Remote tag points to the intended commit
  • GitHub Release exists for the tag
  • Release title exactly equals the tag
  • Release body exactly equals the changelog section body
  • Draft/prerelease/latest state is correct
  • Required native/binary assets exist
  • Publish workflows succeeded
  • Working tree has no accidental release-created changes
  • Downstream dogfood projects no longer use temporary local path dependencies

Use gh release view "$tag" --json name,tagName,body,isDraft,isPrerelease and compare the returned body directly with the prepared notes file. If they differ, repair the release before reporting completion.

Final response

Report only concrete outcomes:

  • Version/package published
  • Documentation published
  • Tag pushed
  • GitHub Release created or repaired
  • Relevant release checks completed
  • Any downstream version update
  • Clean/synchronized repository state

Do not reproduce release-note boilerplate or add promotional/footer links unless the user asks.

Frequently asked questions

What does the Package Release AI skill do?

Prepare, publish, repair, or verify package releases and changelogs. Always use for package/version release, publish, ship, tag, changelog, release notes, registry publication, documentation publication, or GitHub Release work. Distinguish package releases from application deployments before acting.

Why use Package Release on TypingMind?

Because you install it once and use it with any model. Package Release 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 Package Release in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/dannote/dot-pi/tree/master/skills/package-release. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Package Release?

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 Package Release?

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

Is the Package Release AI skill free?

Yes. It is published on GitHub by dannote under the MIT 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 👇