Release Manager logo

Release Manager

Community
luongnv89
release-manager

Manage software releases end-to-end: bump version, generate changelog, tag, push, GitHub release, publish to PyPI/npm. Use when asked to ship, cut a release, or tag a version. Don't use for routine commits or marketplace publishing.

Overview

Publisherluongnv89
Repositoryskills
Skill namerelease-manager
Stars
124
Forks
18
Bundled files
8
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.

  • 8 bundled files

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

  • Open source

    Published by luongnv89 on GitHub. Read the source before you install it.

Installation

Install the Release Manager 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/luongnv89/skills.git /tmp/skills
mkdir -p .claude/skills
cp -r /tmp/skills/skills/release-manager .claude/skills/release-manager
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Release Manager 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 Release Manager 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 Release Manager 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.

Release Manager

Automate the entire release lifecycle: version bump, changelog, README update, documentation sync, build, git tag, GitHub release, and publishing to PyPI/npm.

Architecture (summary)

Main agent orchestrates; heavy steps (scan files, generate changelog, update docs, update landing page) run as parallel subagents to keep context clean. See references/orchestration.md for the full architecture diagram, repo-sync rules, and subagent spawn details. If the Agent tool is unavailable, run the same logic inline.

Overview

A release typically involves these steps in order. Walk through each, confirming with the user before changes. Step 1 can short-circuit the rest: if the project already ships a release tool (.changeset, .releaserc, semantic-release, lerna.json), defer to it and skip steps 2-10.

  1. Pre-flight checks — clean working tree, synced with remote
  2. Determine version — analyze changes, suggest semver bump
  3. Bump version numbers(subagent) scan and propose version changes
  4. Generate changelog / release notes(subagent) from git history and PRs
  5. Update README(subagent, combined with docs) version badges, changelog entries
  6. Update documentation(subagent) sync all project docs 6b. Update landing page(subagent, runs in parallel with 3-6) if the project ships a landing page, refresh its version display, download/install CTA, "What's New", and feature highlights. Clean no-op when there's no landing page. Peer to the docs step, not part of it.
  7. Build — run the project's build step if one exists
  8. Commit, tag, push — create the release commit and tag
  9. GitHub Release — publish on GitHub with release notes
  10. Publish to registries — publish to PyPI and/or npm

Prerequisites

  • Clean working tree (or user-approved stash)
  • Local branch synced with origin (see references/orchestration.md for sync commands)
  • For publishing: PyPI/npm credentials configured; for GitHub release: gh CLI authenticated

Repo Sync Before Edits (mandatory)

Before creating/updating/deleting files in an existing repository, sync the current branch with remote:

bash
branch="$(git rev-parse --abbrev-ref HEAD)"
git fetch origin
git pull --rebase origin "$branch"

If the working tree is not clean, stash first, sync, then restore:

bash
git stash push -u -m "pre-sync"
branch="$(git rev-parse --abbrev-ref HEAD)"
git fetch origin && git pull --rebase origin "$branch"
git stash pop

If origin is missing, pull is unavailable, or rebase/stash conflicts occur, stop and ask the user before continuing.


Step 1: Pre-flight Checks (inline)

If an existing release tool (.changeset, .releaserc, semantic-release config, etc.) is detected, defer to it instead of running the manual steps below — see "Check for existing release tools" further down in this step.

Verify the repo is in a clean state:

bash
git status --porcelain
git rev-parse --abbrev-ref HEAD
git fetch origin
git status -sb

If there are uncommitted changes, ask the user whether to stash, commit, or abort. Never silently discard work.

Check for existing release tools

bash
grep -E '"(release|version|publish)"' package.json 2>/dev/null
ls .releaserc* .changeset/ .versionrc* lerna.json 2>/dev/null

If found, tell the user: "This project uses <tool>. I'll run its release command instead of manual steps." and defer to that tool.


Step 2: Determine Version (inline)

Analyze changes since the last tag:

bash
git tag --sort=-creatordate | head -10
git log $(git describe --tags --abbrev=0 2>/dev/null || echo "HEAD~50")..HEAD --oneline --no-merges

Recommend a bump using conventional commits:

  • MAJOR — any BREAKING: or !: commits
  • MINOR — any feat: (no breaking)
  • PATCH — only fix:, docs:, chore:, refactor:, etc.

Present: "Based on N features, M fixes, K breaking changes since vX.Y.Z, I recommend vA.B.C. Confirm or override?" When in doubt, lean MINOR over PATCH.


Steps 3-6: Parallel Subagent Execution

Once the user confirms the version, spawn the four subagents (version-bumper, changelog-generator, docs-updater, landing-page-updater) in the same turn for parallel execution. The landing-page-updater first checks whether a landing page exists and skips cleanly if not; it edits only release-narrative content, leaving raw version-string bumps to the version-bumper so the two never touch the same line. After they finish, optionally spawn release-reviewer for a quality check (it also flags any cross-agent collision). See references/orchestration.md for the full workspace setup, agent spawn parameters, result collection, apply order, and apply-changes workflow.


Step 7: Build (inline)

Detect the build command:

bash
[ -f package.json ] && grep -q '"build"' package.json && echo "npm run build"
[ -f Makefile ] && grep -q '^build:' Makefile && echo "make build"
[ -f Cargo.toml ] && echo "cargo build --release"
[ -f pyproject.toml ] && echo "python -m build"

Ask the user before running. If the build fails, stop and help debug — never continue with a broken build. If no build step exists, skip and tell the user.


Step 8: Commit, Tag, Push (inline)

Stage changed files (version bumps, changelog, README, docs) and commit:

bash
git add <specific files that were changed>
git commit -m "chore(release): vX.Y.Z"
git tag -a vX.Y.Z -m "Release vX.Y.Z"

Confirm before pushing (see Acceptance Criteria):

bash
git push origin <branch>
git push origin vX.Y.Z

Step 9: GitHub Release (inline)

If gh CLI is available and the repo is on GitHub:

bash
gh release create vX.Y.Z \
  --title "vX.Y.Z" \
  --notes-file CHANGELOG.md \
  --latest

Append artifact paths (.tar.gz, .zip, binaries, .skill files) at the end of the command if any exist. Share the release URL with the user.


Step 10: Publish to Package Registries (inline)

If the project publishes to PyPI and/or npm, read references/publishing.md for the full workflow (pre-requisites, build, verify, upload, post-publish verification).


Expected Output

A successful release ends with the agent printing this expected output:

Release v2.4.0 complete.

Version bumped: pyproject.toml, package.json (1.3.1 → 2.4.0)
Changelog: CHANGELOG.md updated (8 commits: 3 features, 4 fixes, 1 breaking change)
Git tag: v2.4.0 (annotated) pushed to origin
GitHub release: https://github.com/owner/repo/releases/tag/v2.4.0
Published: PyPI — https://pypi.org/project/mypackage/2.4.0/

Post-release reminders:
- [ ] Announce on Discord
- [ ] Monitor for install issues (pip install mypackage==2.4.0)

Edge Cases

  • No conventional commits — version bump cannot be auto-suggested. Show the raw commit list and ask the user to confirm the semver bump explicitly.
  • No remote configuredgit remote returns nothing. Skip push and GitHub release; offer a local tag only.
  • Already published version — target version exists on PyPI/npm (detected via pip index versions or npm view). Abort the publish step and ask whether to bump again or skip publishing.
  • Build artifacts missing — for projects requiring built artifacts, refuse to publish until Step 7 succeeds.

Acceptance Criteria

  • Version string is bumped consistently in all detected files (e.g., pyproject.toml, package.json, __version__)
  • CHANGELOG.md has a new entry for the release version
  • Annotated git tag created and pushed to origin
  • GitHub release created with notes when gh is available
  • User is asked to confirm before each destructive or visible action (push, publish, GitHub release)
  • Post-release checklist is presented after completion

Step Completion Reports

After each major step, output a status report. The full template and per-step variants live in references/step-reports.md. Quick form:

◆ [Step Name] (step N of M — context)
··································································
  Check 1:    √ pass
  Check 2:    × fail — reason
  Criteria:   √ N/M met
  ____________________________
  Result:     PASS | FAIL | PARTIAL

Post-Release Checklist

Remind the user about common follow-ups:

  • Announce the release (blog, social, Discord, Slack)
  • Bump to next dev version (e.g., X.Y.Z-dev) if the project uses that convention
  • Close the GitHub milestone if one exists
  • Monitor for issues
  • Verify packages install (pip install pkg==X.Y.Z, npm install pkg@X.Y.Z)

Tips

  • Confirmation gate for destructive/visible actions: see Acceptance Criteria
  • For monorepos, handle each package's version independently
  • Respect the existing CHANGELOG format — only add the new entry, don't reformat
  • If a release goes wrong mid-way, help the user roll back: delete the tag locally and remotely, revert the commit

Reference files

  • references/orchestration.md — Architecture, repo-sync rules, parallel subagent workflow
  • references/step-reports.md — Full step-completion report templates
  • references/publishing.md — PyPI / npm publishing workflow
  • agents/version-bumper.md — Subagent prompt for version string changes
  • agents/changelog-generator.md — Subagent prompt for changelog generation
  • agents/docs-updater.md — Subagent prompt for documentation updates
  • agents/landing-page-updater.md — Subagent prompt for landing-page updates (skips if none exists)
  • agents/release-reviewer.md — Subagent prompt for independent quality review

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 Release Manager AI skill do?

Manage software releases end-to-end: bump version, generate changelog, tag, push, GitHub release, publish to PyPI/npm. Use when asked to ship, cut a release, or tag a version. Don't use for routine commits or marketplace publishing.

Why use Release Manager on TypingMind?

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

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/luongnv89/skills/tree/main/skills/release-manager. 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 Release Manager?

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 Release Manager?

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

Is the Release Manager AI skill free?

Yes. It is published on GitHub by luongnv89 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 👇