agentleFS
SaehwanParkSaehwanPark/agent-skills-archive/skills/release-preparer/SKILL.md

release-preparer

Skill1 starsChanged 5 months agoMITView on GitHub

Use when converting a private development repository into a clean public release, isolating legacy code, adding repeatable release automation, or preparing Python/PyPI, Rust/Cargo, and Mojo/Pixi packages for publication.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 9 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Treat secret removal and history rewriting as separate security-sensitive work.

    line 15
  2. Distinctive:

    Prefer existing build and release tooling over a parallel release system.

    line 17
  3. Distinctive:

    Use dry runs, temporary outputs, and explicit destination checks before copying, deleting, tagging, publishing, or creating external …

    line 19
  4. Distinctive:

    Always read public repository guidance for public-release curation.

    line 24
  5. Distinctive:

    Do not modernize unrelated implementation code.

    line 56
  6. Distinctive:

    Do not publish secrets, private data, internal-only docs, or unclear-license material.

    line 57
  7. Distinctive:

    Do not delete legacy code merely because it is excluded from the release.

    line 58
  8. Distinctive:

    Do not claim reproducibility without running the documented build path.

    line 59
  9. Distinctive:

    Stop when ownership or licensing is unclear, secrets may remain in history, the destination could overwrite existing work, package …

    line 63

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Release Preparer

Prepare a public release without rewriting the working project or publishing anything without explicit authorization.

Operating rules

  • Inspect repository instructions, status, history, ignore rules, licenses, secrets, generated artifacts, package metadata, tests, and release configuration first.
  • Treat secret removal and history rewriting as separate security-sensitive work; deleting a current file does not remove it from history.
  • Prefer existing build and release tooling over a parallel release system.
  • Preserve package identity and compatibility unless the release request changes them.
  • Use dry runs, temporary outputs, and explicit destination checks before copying, deleting, tagging, publishing, or creating external resources.

Reference routing

Always read public repository guidance for public-release curation. Then read only the relevant ecosystem guidance:

  • Python and PyPI
  • Rust and Cargo
  • Mojo and Pixi

For multi-package repositories, load each applicable guide and keep shared release policy at the repository boundary rather than duplicating it in every package.

Workflow

  1. Define release source, destination, included packages, excluded private/legacy material, version, and required publication targets.
  2. Inventory tracked and untracked content, credentials, licenses, generated files, package manifests, tests, and current automation.
  3. Classify each material item as publish, generate, ignore, archive separately, or stop for user direction.
  4. Make the smallest changes needed for a reproducible build and clean source artifact.
  5. Validate metadata, package contents, documentation links, licenses, builds, tests, and clean-environment installation where practical.
  6. Preview release or repository operations before executing authorized external actions.
  7. Report artifacts, commands, omissions, approvals still needed, and residual risks.

Approval-gated actions

Require explicit authorization for destructive cleanup, history rewriting, creating or replacing a public repository, pushing, tagging, creating releases, and publishing to a registry. Confirm the exact target and version immediately before an irreversible action.

Non-goals

  • Do not modernize unrelated implementation code.
  • Do not publish secrets, private data, internal-only docs, or unclear-license material.
  • Do not delete legacy code merely because it is excluded from the release.
  • Do not claim reproducibility without running the documented build path.

Stop conditions

Stop when ownership or licensing is unclear, secrets may remain in history, the destination could overwrite existing work, package identity/version conflicts exist, validation cannot be reproduced, or required publication authority is absent.

Expected output

Report release boundaries, files or automation changed, validation and package inspection results, external actions taken, deferred approvals, and remaining security or compatibility risks.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.