skills/agr-release/SKILL.md
Release process for the agr package. Handles version bumping (major/minor/patch/beta), changelog updates, pre-release quality checks, git tagging, and monitoring the GitHub Actions publish pipeline. Use this skill whenever the user wants to cut a release, bump the version, publish to PyPI, or asks about the release process — even if they just say "let's ship it" or "time for a new version".
npx skillsauth add kasperjunge/agent-resources agr-releaseInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
This skill walks through the full release process for the agr package. The release is
tag-driven: pushing a vX.Y.Z tag triggers the GitHub Actions pipeline that runs quality
checks, builds the package, publishes to PyPI, and creates a GitHub Release.
Your job is to prepare everything so that when the tag is pushed, the pipeline succeeds on the first try.
Verify the preconditions. If any fail, stop and tell the user.
git status should show no uncommitted changesmain branch — releases should only come from maingit pull to make sure you're not behindAsk the user what kind of release this is:
If the user already said what type they want, don't ask again.
Before touching any files, understand what's being released.
# See all commits since the last release tag
git log $(git describe --tags --abbrev=0)..HEAD --oneline
Also check the [Unreleased] section in CHANGELOG.md — it may already have entries. Cross-reference with the git log to make sure nothing is missing. If there are commits that aren't reflected in the changelog, add them.
Group changes into the standard Keep a Changelog categories:
Run all three locally before proceeding. These are the same checks the CI pipeline runs, so catching failures here saves a round-trip.
uv run ruff check .
uv run ruff format --check .
uv run pytest -m "not e2e and not network and not slow"
uv run ty check
If anything fails, fix it before continuing. The release commit should pass CI cleanly.
Not every release needs doc changes — use judgement. Docs updates are warranted when:
The docs to consider:
README.md — the primary entry point, should reflect current capabilitiesdocs/docs/reference.md — CLI command referencedocs/docs/index.md — landing page / getting starteddocs/docs/ as relevant (sdk.md, configuration.md, etc.)skills/ — if any exist and are affected by the changesIf nothing user-facing changed (internal refactors, test improvements, dependency bumps), skip this step and move on.
The version lives in pyproject.toml (the single source of truth — importlib.metadata picks it up at runtime via agr/__init__.py):
pyproject.toml line 7: version = "X.Y.Z"Calculate the new version based on the current version and the release type the user chose.
For beta releases, append b1 (or increment the beta number if one already exists):
0.7.10 → 0.7.11b1 (first beta of next patch)0.7.11b1 → 0.7.11b2 (next beta)0.7.11b2 → 0.7.11 (promote beta to stable)In CHANGELOG.md:
## [Unreleased] with ## [X.Y.Z] - YYYY-MM-DD (today's date)## [Unreleased] section at the topThe changelog format matters because the GitHub Actions pipeline extracts the version's section to use as release notes. Malformed entries = bad release notes.
# Stage the changed files
git add pyproject.toml agr/__init__.py CHANGELOG.md
# Plus any docs files you updated
# Commit
git commit -m "release: vX.Y.Z"
# Tag
git tag vX.Y.Z
# Push commit and tag
git push origin main
git push origin vX.Y.Z
Wait for the user to confirm before pushing. Show them a summary of what will be pushed:
After pushing the tag, monitor the GitHub Actions pipeline:
# Watch the workflow run
gh run list --workflow=publish.yml --limit=1
gh run watch $(gh run list --workflow=publish.yml --limit=1 --json databaseId -q '.[0].databaseId')
The pipeline has four stages:
uv build + verifyIf any stage fails, read the logs and help the user fix it:
gh run view <run-id> --log-failed
Common failure modes:
Once the pipeline succeeds, confirm:
# Check PyPI (may take a minute to propagate)
pip index versions agr
# Check the GitHub release exists
gh release view vX.Y.Z
Tell the user the release is live and share the links:
gh release viewIf the pipeline fails and you need to retry:
git tag -d vX.Y.Z && git push origin :refs/tags/vX.Y.ZThis is destructive — confirm with the user before deleting tags.
tools
Debrief an AI agent skill (SKILL.md) after using it — capture session feedback or a retrospective and fold it back into the skill. Use whenever the user says: "debrief the X skill", "let's debrief X", "retrospective on X", "feedback on X skill", "improve the X skill", "update the X skill", "let's revise X based on what we just did", "the X skill should also handle Y", "X didn't trigger when it should have", or otherwise wants to capture lessons from a session back into the skill that drove it. Handles in-repo skills (under skills/) by editing the source, committing, and re-installing via `agr upgrade`. Handles remote/upstream skills by offering to fork them in-repo or to file a GitHub issue via `gh`. Do NOT use for greenfield skill authoring (a separate concern — see `anthropics/skills/skill-creator`) or for installing / syncing / removing skills (use the `agr` CLI directly).
tools
Install, share, sync, and create AI agent skills across coding tools (Claude Code, Cursor, Codex, OpenCode, Copilot, Antigravity) using the agr CLI. Use whenever the user mentions agr, agr.toml, agr.lock, agrx, or asks to: add a skill ("install the pdf skill", "agr add ..."), sync agent resources across tools, share skills with their team, scaffold a new SKILL.md, run a skill ephemerally (agrx), set up a repo to manage AI agent dependencies, or configure tools/sources/instruction-syncing. Also use whenever an agr.toml or agr.lock is present in the project and the user is doing resource-management work.
testing
Create, edit, improve, or audit AgentSkills. Use when creating a new skill from scratch or when asked to improve, review, audit, tidy up, or clean up an existing skill or SKILL.md file. Also use when editing or restructuring a skill directory (moving files to references/ or scripts/, removing stale content, validating against the AgentSkills spec). Triggers on phrases like "create a skill", "author a skill", "tidy up a skill", "improve this skill", "review the skill", "clean up the skill", "audit the skill".
testing
Host security hardening and risk-tolerance configuration for OpenClaw deployments. Use when a user asks for security audits, firewall/SSH/update hardening, risk posture, exposure review, OpenClaw cron scheduling for periodic checks, or version status checks on a machine running OpenClaw (laptop, workstation, Pi, VPS).