English · 简体中文
A Codex skill for turning a working repository into a launch-ready open-source project.
It helps you sharpen positioning, improve the README and quickstart, plan convincing demo assets, and prepare Chinese/English release copy—without fake growth or noisy marketing.
30-second walkthrough: audit, Markdown report, launch package, benchmark map, and bilingual copy.
Audit → Position → Benchmark → Package → Demo → Launch
Use GitHub Launch Studio before:
- publishing a project on GitHub for the first time;
- preparing a major release;
- improving a repository that works but does not clearly communicate its value;
- creating a demo GIF, screenshots, social card, or release-page story;
- launching a Chinese-built project to an English-speaking audience.
Audit this repository and tell me whether it is ready to launch.
Rewrite the README for an international GitHub launch.
Create a v1.0 release plan with GitHub, X, V2EX, and 即刻 posts.
Design a 30-second demo GIF storyboard for this CLI.
Give me a prioritized launch checklist for next week.
- Evidence-based launch-readiness scores across clarity, activation, proof, trust, distribution, and maintenance.
- A concise positioning statement, target user, five-minute promise, differentiator, and non-goals.
- A grounded benchmark map across README builders, release automation, changelog tools, analytics, and agent workflow kits.
- README and quickstart recommendations with a first-time-user test.
- Demo asset briefs for screenshots, GIFs, videos, and social cards.
- GitHub release notes and channel-specific Chinese/English announcement copy.
- A publishing checklist that keeps secrets, permissions, branch protection, CI, and feedback triage in scope.
GitHub Launch Studio is complementary to release automation, not a replacement for it:
| Tool type | Strong at | GitHub Launch Studio adds |
|---|---|---|
| README editors | Filling and arranging README sections | Positioning, evidence, quickstart quality, and launch narrative |
| Profile README generators | Personal GitHub profile components | Project launch packaging rather than a personal profile |
| semantic-release / release-please / git-cliff / Changesets | Versioning, changelogs, release PRs, and package publishing | Human-facing launch story, demo plan, bilingual copy, and channel plan |
| Star History | Post-launch star trends | Pre-launch activation and distribution work |
| Agent spec/workflow kits | Structured development process | A specialized launch-growth workflow |
See the full benchmark map / 中文对标说明. Metrics are observed on a stated date and can be refreshed with the bundled script.
After the repository is published:
git clone https://github.com/MJorgin/github-launch-studio.git \
~/.codex/skills/github-launch-studioFor local development, copy or symlink this folder into your Codex skills directory, then restart or reload Codex.
The bundled audit script works offline and never prints credential file contents. The scan itself is read-only; it writes a file only when --output is provided.
python3 ~/.codex/skills/github-launch-studio/scripts/repo_audit.py /path/to/repository --prettyGenerate a shareable Markdown report:
python3 scripts/repo_audit.py /path/to/repository \
--format markdown \
--output /path/to/repository/LAUNCH.mdIt checks signals such as:
- README, LICENSE, docs, examples, lockfiles, manifests, and workflows;
- quickstart headings, code blocks, images, badges, and broken relative links;
- issue templates, contributor paths, security policy, and roadmap/changelog signals;
- likely secret-bearing filenames;
- simple 0-5 launch-readiness scores and prioritized findings.
The scores are heuristics for a first pass. The skill should still inspect the repository semantically before recommending product positioning or copy.
- 30-second demo video / animated GIF — the rendered walkthrough included in the README.
- 30-second demo storyboard — recording plan, captions, and pre-flight checks.
- Draft launch posts — X, V2EX, and 即刻 drafts for review before posting.
python3 scripts/benchmark_github.py \
--config benchmarks/representative-tools.json \
--format markdownThe benchmark helper uses the public GitHub API. Set GITHUB_TOKEN to raise the rate limit; the script never prints token values.
The repository includes a rendered assets/social-card.png. To regenerate it locally:
python3 -m pip install pillow
python3 scripts/render_social_card.py --output assets/social-card.pngThe deterministic demo requires Pillow and ffmpeg:
python3 -m pip install pillow
python3 scripts/render_launch_demo.pyIt writes assets/launch-demo.mp4 and assets/launch-demo.gif without recording the desktop or exposing local terminal paths.
- Audit the repository using real file and Git metadata.
- Define the target user, five-minute outcome, proof, and strongest alternative.
- Improve activation surfaces: hero, quickstart, example, limits, and support path.
- Plan the smallest demo asset that proves the value.
- Prepare bilingual release notes and channel-specific posts.
- Verify the clean-install path and publish only after explicit approval.
Read SKILL.md for the full operating rules and the references directory for detailed playbooks.
- The audit script does not use the network.
- Drafts and local edits are separated from external publication.
- Creating GitHub releases, changing repository settings, posting on social platforms, commenting in communities, or sending DMs requires separate authorization.
- The skill explicitly avoids bought stars, fake testimonials, undisclosed paid promotion, mass outreach, scraped contact lists, or fabricated metrics.
- Optional Markdown/link report rendering.
- More repository-language heuristics and clean-install probes.
- Example launch teardowns from high-quality open-source projects.
- Optional bilingual release-note templates for common ecosystems.
See CONTRIBUTING.md and SECURITY.md.
- preflight-decks — concept-first director for HTML pitch/talk decks: a hard concept gate, three real visual directions, and dual-viewport verification. From the same author.