release-all
Skill80 starsChanged 4 months agoMITView on GitHub
Publish the complete Spool release train: synchronized versions, desktop artifacts, npm packages, and the production web deployment.
We split this skill into the rules it gives an agent
9 rules. 0 are common, 3 are distinctive, and 6 are very specific to this project.
0
Common
Most of the 4 edited versions we read keep these.
3
Distinctive
This author's own choices. Worth a look.
6
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 10
Treat a Spool release as a release train: one version must reach every target defined by the repository.
- Distinctive: line 28
Run the preflight checks before changing files:
- Very specific: line 48
every non-automatic secret referenced by the release workflow appears in gh secret list
- Very specific: line 51
If the repository instead shows evidence of an interrupted release—such as a release: v… commit or matching local tag—read RECOVERY.md and …
- Distinctive: line 65
Use a timeout long enough for GitHub Actions to build, sign, notarize, publish, and return.
- Very specific: line 67
If any command fails after launch begins, continue the same version through RECOVERY.md.
- Distinctive: line 85
Verify against the target map from step 1:
- Very specific: line 87
Every manifest changed by release.sh contains the new version.
- Very specific: line 91
Every npm publish target resolves at the new version with npm view <name>@<version> version.
“Common” is measured: kept by at least 6 in 10 of the 4 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.
Files
Release All
Treat a Spool release as a release train: one version must reach every target defined by the repository. scripts/release.sh, .github/workflows/release.yml, and .github/workflows/deploy-web.yml are the sources of truth.
Invoking this skill authorizes the version commit, tag, push, package publication, GitHub release, and production deployment. Proceed without another confirmation after the gates pass.
1. Map the train
From the repository root, read the three source-of-truth files above. Interpret $ARGS as follows:
- empty or
patch→ patch release minor→ minor release- any other value → stop and state the supported values
Derive the current version, next version, tag, manifests changed by the release script, npm publish targets, expected release assets, and required repository secrets. Do not maintain a second hard-coded target list in this skill.
This step is complete when every derived target has a concrete verification check for the next version.
2. Gate the departure
Run the preflight checks before changing files:
git fetch origin main --tags
git branch --show-current
git status --porcelain
git rev-parse HEAD
git rev-parse origin/main
command -v gh
command -v jq
gh auth status
gh secret list --json name --jq '.[].name'Require all of these conditions:
- the branch is
main - the worktree is clean
- local
HEADequalsorigin/main ghis authenticated andjqis available- every non-automatic secret referenced by the release workflow appears in
gh secret list - the predicted tag is absent locally and on
origin
If the repository instead shows evidence of an interrupted release—such as a release: v… commit or matching local tag—read RECOVERY.md and resume that version. For any other failed gate, stop with the exact condition and one remediation.
This step is complete only when every gate passes or a verified interrupted release has entered recovery.
3. Launch once
Record the predicted version and tag, then run exactly one command:
./scripts/release.sh --patch
# or
./scripts/release.sh --minorUse a timeout long enough for GitHub Actions to build, sign, notarize, publish, and return. After it exits, record the release commit from git rev-parse HEAD.
If any command fails after launch begins, continue the same version through RECOVERY.md. release.sh runs once per release attempt because another invocation would bump a second version.
This step is complete when the exact-tag Release workflow concludes successfully.
4. Wait for production web
The release commit's push to main starts Deploy Web independently. Poll for the run whose headSha equals the recorded release commit, then watch it:
gh run list --workflow=deploy-web.yml --commit "$RELEASE_COMMIT" \
--limit 1 --json databaseId,headSha,status,conclusion,url
gh run watch "$WEB_RUN_ID" --exit-statusPoll for up to 60 seconds before treating a missing run as a recovery case. This step is complete when that exact-commit deployment succeeds.
5. Verify every carriage
Verify against the target map from step 1:
- Every manifest changed by
release.shcontains the new version. origin/maincontains the release commit and the remote tag resolves to it.- The exact-tag Release workflow succeeded.
- The GitHub release exists and contains every artifact class uploaded by the workflow.
- Every npm publish target resolves at the new version with
npm view <name>@<version> version. Retry registry reads for up to two minutes. - Deploy Web succeeded for the release commit.
The release is complete only when all six checks pass. Report a compact table with the version/tag, GitHub release URL, Release workflow URL, npm packages, and Deploy Web URL. If a check remains red, report which carriages are already published and the exact resume action; never describe a partial release as complete.
Other versions4 projects use a version of this.
Word for word · 0
None found.
Edited · 2
- spool-lab/spool★ 59285% same
- paperboytm/spool★ 59085% same
Wrote their own version · 2
- duobaseio/forui★ 2.4ksame name
- forus-labs/forui★ 2.4ksame name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
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
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.

