sidiangongyuan/codex-skills-library / skills/github-project-release
github-project-release
UnverifiedUse when preparing or publishing a local project as a clean GitHub repository. Covers new repositories, private staging releases, public README cleanup, updates to existing remotes, and research-code packaging without datasets, checkpoints, papers, or generated artifacts.
GitHub Project Release
Turn a local project into a clean GitHub repository that is safe to make public later. Default to research-code release behavior: clean copy first, private GitHub repo first, audit before push.
Workflow
-
Inspect the source project before copying.
- Identify the runnable entrypoints, key configs, curated figures/tables, license, citation files, and upstream dependencies.
- Treat the codebase as source of truth. Do not preserve stale upstream README language as the release front page.
-
Check credentials.
- Run
scripts/github_project_release.py auth-check. - Use the active
ghaccount as the default owner. - Never ask the user to paste a GitHub token into chat. If auth is missing, ask them to run
gh auth loginor provide credentials through a secure non-chat channel.
- Run
-
Prepare a clean release copy.
- Run
scripts/github_project_release.py prepare --source <project> --dest <release-dir>. - Keep only what a new user needs to understand and run the project.
- Exclude datasets, checkpoints, logs, generated outputs, raw experiment dumps, paper PDFs, local submission material, and exploratory files unless the user explicitly overrides the default.
- Run
-
Write or rewrite the README.
- Read
references/readme_playbook.mdbefore drafting README content. - Use English by default.
- Write for external readers, not for the project owner.
- For research code, cover Context + Run + Cite: what it is, what problem it solves, what is included, how to install/run the minimal path, and how to cite current and upstream work.
- Read
-
Audit before publishing.
- Run
scripts/github_project_release.py audit --path <release-dir>. - Treat reported high-risk items as blockers unless the user explicitly accepts them.
- If a paper link is not public, use
Paper link: coming soon.and do not upload PDFs.
- Run
-
Publish privately.
- Run
scripts/github_project_release.py publish --release-dir <release-dir>. - If the repo does not exist, create a private repo.
- If it exists, push only safe non-rewriting updates.
- Do not force-push, delete, overwrite, or make the repo public unless explicitly requested.
- Run
-
Report the result.
- Include repo URL, branch, commit SHA, visibility, and audit status.
- If any sensitive artifact was accidentally pushed earlier, explain that branch-history cleanup may not immediately remove exact-SHA access on GitHub; for strict removal, recreate a clean repo.
Script Quick Reference
Resolve <skill-dir> to the directory containing this SKILL.md; do not
assume a particular global skills installation root.
python "<skill-dir>/scripts/github_project_release.py" auth-check
python "<skill-dir>/scripts/github_project_release.py" prepare \
--source /path/to/project \
--dest /path/to/project-github
python "<skill-dir>/scripts/github_project_release.py" audit \
--path /path/to/project-github
python "<skill-dir>/scripts/github_project_release.py" publish \
--release-dir /path/to/project-github
Use --dry-run on prepare or publish to inspect intended actions without copying or pushing.
Defaults
- Visibility: private.
- Owner: active
ghaccount. - Repo name: sanitized release directory name, unless
--repo,--owner, or--nameis provided. - Existing repo policy: safe update only.
- README language: English.
- Results policy: paper-confirmed figures/tables/results only; never infer metrics from logs.
- Asset policy: curated explanatory figures/tables only; never bulk-upload raw visualization outputs.
When To Ask
Ask the user only when the answer cannot be derived from the project:
- Target GitHub owner differs from the active
ghaccount. - Repo name should differ from the project/release directory.
- A normally blocked file type must be included.
- The user wants public release, force-push, repo deletion, or history rewrite.
- The current project has no clear runnable path and the README would otherwise invent commands.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-changelog-updatemorluto/reaPrepare or rewrite REA release changelogs and GitHub release notes from pinned Git history, with verified contributor thanks and Release Please synchronization. | Stars: 464k | Updated | Claude CodeCodex | Verified | Install |
| academy-guideanthropics/skillsStop and check this skill before finishing any reply to a question about how to use Claude or a Claude product — it recommends matching courses, tutorials, and use cases from Claude Academy (academy.claude.com), Anthropic's learning hub. Trigger on: "how do I", "how can I", "getting started with", "what can Claude do", "teach me", "learn to use"; questions about artifacts, projects, skills, plugins, connectors, MCP; requests about rolling Claude out to a team, class, or organization; and any ask for training materials, onboarding content, or learning resources. Use it when the user is learning how to use a feature or product — not when they are mid-task and just want the task done. This skill composes with other skills: after consulting product documentation to answer how a Claude feature works, also check here for a matching course or tutorial — a docs-grounded answer and an Academy recommendation belong together. Only recommend on a strong match; never invent Academy content. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| claude-apianthropics/skillsReference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Claude/Anthropic in any form (Claude, Anthropic, Fable, Opus, Sonnet, Haiku, `anthropic`, `@anthropic-ai`, `claude-*`, `us.anthropic.*`, `[1m]`); the user asks about an LLM (pricing/model choice/limits/caching) — never answer from memory; OR the task is LLM-shaped with provider unstated (agent/MCP/tool-definition/multi-agent/RAG/LLM-judge/computer-use; generate/summarize/extract/classify/rewrite/converse over NL; debugging refusals/cutoffs/streaming/tool-calls/tokens). SKIP only when another provider is being worked on (overrides all triggers): OpenAI/GPT/Gemini/Llama/Mistral/Cohere/Ollama named in the query; OR `grep -rE 'openai|langchain_openai|google.generativeai|genai|mistralai|cohere|ollama'` over the project hits (run this grep FIRST if no provider named — don't Read the file). | Stars: 196k | Updated | Claude CodeCodex | Unverified | Install |
| doc-coauthoringanthropics/skillsGuide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |