agentleFS
zajcazajca/pohunek/.agents/skills/release/SKILL.md

release

Skill8 starsChanged 10 days agoMITView on GitHub

Cut a pohunek release with scripts/release — bump the workspace version, tag vX.Y.Z, push, and verify the GitHub Actions Release workflow builds and publishes the glibc + MUSL x86_64 binaries. Use when the user asks to cut or publish a new version.

193 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use create-agent-adapter, changelog-style and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • MicrosoftSkill

    create-agent-adapter

    Technical guide for creating a new Paperclip agent adapter. Use when building a new adapter package, adding support for a new AI coding tool (e.g. a new CLI agent, API-based agent, or custom process), or when modifying the adapter system. Covers the required interfaces, module structure, registration points, and conventions derived from the existing claude-local and codex-local adapters.

    Also in 5 projects that use this

  • GrinvSkill

    changelog-style

    Turn a commit range into CHANGELOG.md entries following common-changelog style. Use when finishing a PR or cutting a release, to write/fix the [Unreleased] section's entries — moving [Unreleased] under a new dated version heading is a separate, automated step (see the release skill / scripts/sync-version.ts), not this skill's job.

    Also in 4 projects that use this

  • GrinvSkill

    docs-consistency-check

    Check README/manifest.json/server.json/CHANGELOG.md/AGENTS.md and docs/*.md for drift against the actual registered tools and source. Use after adding, renaming, or removing a tool, or as part of a live-audit pass.

    Also in 4 projects that use this

  • Microsoft AzureSkill

    component-release

    Prepares a verifiable Agent Landing Zone UI release. Use for release branches, VERSION, changelog entries, tags, release notes, and compatibility evidence; never publish without explicit human approval.

    Also in 4 projects that use this

  • Microsoft AzureSkill

    documentation-consistency

    Keeps AI Landing Zone contributor, user, and operator documentation aligned with shipped Bicep behavior. Use for parameters, defaults, outputs, topology, deployment flow, operations, or releases.

    Also in 8 projects that use this

  • rivet-devSkill

    launch-graphics

    Generate agentOS marketing/launch graphics (launch & social heroes, code-snippet cards) from the committed generators in docs-internal/graphics. Use when asked to make, update, or render an agentOS launch image, social image, "agentOS Apps" graphic, or technical code-snippet image.

    Also in 3 projects that use this

  • +3 more from Grinv

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 3 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 183 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Cut the release: Run the real thing:

    line 58
  2. Very specific:

    Publish the notes: When a notes file was needed, attach it to the GitHub Release once the workflow has created it (gh release edit "vX.Y.Z" …

    line 83
  3. Distinctive:

    Never publish from a red commit.

    line 120
  4. Very specific:

    Do not hand-edit versions or push a tag outside scripts/release.

    line 122
  5. Distinctive:

    Verify the release actually published.

    line 123

“Common” is measured: kept by at least 6 in 10 of the 183 edited versions of this file we could read. “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 — cut and publish a release

Wraps scripts/release and the Release GitHub Actions workflow (.github/workflows/release.yml). Do the version bump through the script — never hand-edit Cargo.toml/Cargo.lock versions or hand-craft the tag.

Preconditions (the script enforces these; check them first)

  • On the release branch — main by default (override with POHUNEK_RELEASE_BRANCH). The script aborts if you are on another branch.
  • Working tree is clean. Commit or stash first; the script aborts on a dirty tree.
  • The gates pass on the commit being released. The Release workflow re-runs fmt/clippy/test (the nextest shards CI runs, plus doctests) + docs check on the tag, so a red commit publishes nothing — run the gates skill before tagging to fail fast locally.
  • The target tag does not already exist.

Steps

  1. Compare the compatibility constants. The script diffs these constants between the previous tag and HEAD: STORE_SCHEMA_VERSION (crates/daemon/src/store/), PROTOCOL_VERSION and MIN_PROTOCOL_VERSION (crates/protocol/src/version.rs), the worker protocol CURRENT_VERSION and PREVIOUS_VERSION (crates/worker-protocol/src/version.rs), WORKER_JOURNAL_SCHEMA_VERSION (crates/daemon/src/runtime/lifecycle.rs), and EXPECTED_INTEGRATION_VERSION (crates/protocol/src/integration.rs). Each one that changed alters the upgrade contract in .agents/instructions/project.md ("Upgrade window"), so the release needs a notes file with one line naming each changed constant and what an operator must do. Write it before the dry run and pass it as --notes FILE; the script aborts when a changed constant has no line.
  2. Confirm the bump. Determine patch, minor, major, or an explicit X.Y.Z from the user's request ("minor release" → minor).
  3. Dry-run first. Show what will happen without changing anything:

    scripts/release <patch|minor|major|X.Y.Z> [--notes FILE] --dry-run

    This prints the current and next version, the tag, the knowledge-log fragments the release would file, the folded CHANGELOG.md section that becomes the release body, and whether the release notes cover any changed compatibility constants. It also describes the commit, evidence checks, and pushes it would perform. Confirm that matches intent.

  4. Cut the release. Run the real thing:

    scripts/release <patch|minor|major|X.Y.Z> [--notes FILE]

    The script files the docs/knowledge/log.d/ fragments under ## X.Y.Z in docs/knowledge/log.md and removes them (cargo xtask docs log fold), folds the changelog.d/ fragments into a ## X.Y.Z - <UTC date> section of the committed CHANGELOG.md and removes them (scripts/changelog fold), bumps version in [workspace.package] in Cargo.toml, refreshes Cargo.lock, commits all of it as Release vX.Y.Z, and requires successful main and nightly CI evidence for that commit before creating an annotated tag vX.Y.Z. Evidence is a run on the commit itself or a workflow_dispatch run that tested the same git tree; scripts/ci-evidence prints which run matched and whether by sha or by tree. By default it pushes the release branch before waiting for evidence and pushes the tag after verification. With --no-push, it checks only already-existing evidence and leaves the commit and tag local; it does not dispatch CI or push either ref. After local tagging, the script reminds you to push the release branch and tag manually. If the evidence check fails or times out, resolve the CI issue and rerun with the same explicit X.Y.Z; the workspace is already bumped, so do not request another relative bump such as patch.

  5. Publish the notes. When a notes file was needed, attach it to the GitHub Release once the workflow has created it (gh release edit "vX.Y.Z" --notes-file FILE) and confirm with gh release view "vX.Y.Z".
  6. Record the release. Via the github-workflow skill, comment the version/tag and release URL on any issue whose delivery this release completes (or opens a dedicated tracking issue beforehand when the release itself is planned work), and update the project status accordingly. Verify the writes from the API response.
  7. Verify the Release workflow — do not trust, confirm. Pushing the vX.Y.Z tag triggers .github/workflows/release.yml, which runs the fmt/clippy/test gate + docs-gate, then builds pohunek and pohunekd for both x86_64-unknown-linux-gnu (dynamic glibc, primary) and x86_64-unknown-linux-musl (fully static, runs on any x86_64 Linux), packages per-component tarballs with sha256 checksums in read-only build jobs (release-build.yml), builds the official runtime package archives, runs the compatibility rows, assembles and signs the runtime catalog and rebuilds the daemon archives with a runtime/ directory (release-evidence.yml), runs the network-isolated archive smoke, and only then the single publish job (the only job with a write token, no checkout) verifies the complete inventory, creates a draft release, uploads and checks it, and publishes it. The TypeScript SDK tarballs pohunek-ts-{protocol,sdk,testkit}-X.Y.Z.tgz, each with a .sha256, are part of the same inventory. Watch it to completion:

    gh run watch "$(gh run list --workflow=release.yml --limit 1 --json databaseId --jq '.[0].databaseId')"
    gh release view "vX.Y.Z"   # confirm the tarballs + .sha256 (incl. the three pohunek-ts-*.tgz) are attached
  8. Report. State the published version/tag, the workflow conclusion (success/failure with the failing job if any), and the attached artifacts. If the workflow failed, report why with output — a failed gate means no binary was published.

Constraints

  • Never publish from a red commit; the workflow will block it, so catch it locally.
  • Do not hand-edit versions or push a tag outside scripts/release.
  • Verify the release actually published; never report "released" from the tag push alone without checking the workflow and the GitHub Release.

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.