agentleFS
BuilderIOBuilderIO/agent-native/.agents/skills/ship-and-monitor/SKILL.md

ship-and-monitor

Skill7.6k starsChanged 5 months agoMITView on GitHub

Run the normal guarded ship flow, then do requested post-merge checks. Add independent beta behavior checks only when requested or a concrete deployment-specific risk makes local proof insufficient. Use when standard `/ship` needs post-merge beta/live, docs, release, or production verification.

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 6 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use the normal /ship workflow, then add the post-merge monitoring described below.

    line 16
  2. Distinctive:

    Never cancel, re-dispatch, or pin duplicate beta publisher runs.

    line 31
  3. Distinctive:

    When beta verification is requested or warranted by the gate below, use the first successful publish whose source contains your merge.

    line 32
  4. Very specific:

    Use .github/workflows/deploy-production-sites-prebuilt.yml or the targeted promote-netlify-deploy.yml workflow to promote a critical fix …

    line 43
  5. Distinctive:

    When changed auth behavior depends on beta host/cookie settings, provider callback registration, deployed auth configuration, or serverless …

    line 50
  6. Distinctive:

    Check applicable workflows and package publication.

    line 80
  7. Very specific:

    If the task changed docs, verify the docs production workflow and www.agent-native.com separately.

    line 85
  8. Distinctive:

    When beta verification is in scope, correct a beta update already posted in an owned feedback thread if the matching publish fails or the …

    line 94
  9. Distinctive:

    When deployed proof is in scope, keep configured, source-tested, built-runtime, beta-deployed, production-promoted, and observed-live …

    line 99

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 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

Ship and Monitor

Use the normal /ship workflow, then add the post-merge monitoring described below. Read .agents/skills/ship/SKILL.md and follow every step through /new-branch; /ship intentionally ends after merge and branch rotation. A merge-triggered beta publish alone is not a reason for an independent beta browser, URL, or E2E check.

Deployment split

Merges to main trigger .github/workflows/deploy-beta-sites-prebuilt.yml, which builds in GitHub Actions and uploads prebuilt artifacts to the independent Netlify beta sites at beta.*.agent-native.com. Netlify Git-connected auto-builds are disabled, so do not wait for Netlify build queues or deploy-preview checks; verify the Actions run and its built-in per-site smoke checks when beta deployment is in scope. Those checks do not require an extra independent app smoke for routine source changes. Never cancel, re-dispatch, or pin duplicate beta publisher runs; automatic/handoff runs coalesce pending runs, manual runs use run-specific groups, and publishes serialize. When beta verification is requested or warranted by the gate below, use the first successful publish whose source contains your merge. Do not wait for beta solely because the merge triggered a publish. Production promotion is manual. A healthy beta deploy is not proof that production changed, and a production deploy is not expected unless an explicit manual promotion was started for the task. The public docs site is the temporary exception: matching main changes trigger .github/workflows/deploy-docs-production.yml, which publishes www.agent-native.com directly and then disables its Git-connected Netlify builds. There is no beta docs site today.

Use .github/workflows/deploy-production-sites-prebuilt.yml or the targeted promote-netlify-deploy.yml workflow to promote a critical fix and let it manage Netlify lock transitions. Do not manually remove or clear a Netlify lock as a deployment step; clearing one is not the production promotion.

Host-dependent auth gate

When changed auth behavior depends on beta host/cookie settings, provider callback registration, deployed auth configuration, or serverless session behavior, local proof cannot cover that boundary. After the affected beta sites deploy, dispatch both lanes below from main, using the exact affected app ids. Auth screen copy, layout, and other changes that do not touch those boundaries do not need beta E2E by default:

gh workflow run signup-e2e-scheduled.yml --ref main \
  -f apps=<email-signup-affected-apps> \
  -f environments=beta
gh workflow run beta-e2e.yml --ref main \
  -f lane=public+authed \
  -f apps=<affected-beta-apps>

The signup dispatcher selects its apps and environments independently of the browser dispatch's apps and lane inputs. Wait for the signup job's classified output to be exactly success and for the affected browser lane to pass. Complete each touched Google callback in a real browser session as well; the seeded authenticated lane deliberately excludes Google-only Mail and Calendar. A failure, cancellation, inconclusive Mailosaur result, missing beta deploy, or untested provider path stays Open - do not report the auth fix as done.

Post-merge monitoring

After /new-branch:

  1. Confirm the merged PR and merge commit are present in origin/main.
  2. Check applicable workflows and package publication. Wait for the affected beta deploy and exercise its exact path only when beta proof was requested or a concrete beta-specific risk applies (including the auth gate above). Otherwise do not wait solely for beta or add independent beta browser, URL, or E2E checks; local proof and required CI cover ordinary source changes.
  3. If the task changed docs, verify the docs production workflow and www.agent-native.com separately. If the task explicitly included manual promotion of another production site, verify that promotion and its URL separately. Otherwise report those other production sites as intentionally not promoted, not as blocked by Netlify.
  4. Re-read the merged PR for new review or bot feedback. If actionable post-merge feedback or a release/deploy failure appears, fix it on the fresh branch, run the smallest meaningful check, and invoke /ship for the follow-up.
  5. When beta verification is in scope, correct a beta update already posted in an owned feedback thread if the matching publish fails or the affected site is still not current 24 hours after the update. The task running that check owns the correction.

When deployed proof is in scope, keep configured, source-tested, built-runtime, beta-deployed, production-promoted, and observed-live claims separate. A merge or green test alone is not live proof.

Related skills

  • /ship for the normal guarded flow without post-merge monitoring.
  • /ship-now for the fast admin-merge path; it watches the merge commit's workflows itself.
  • /new-branch for the required branch rotation after merge.

Download now (ZIP)

Trending

More talk lately.about the whole repository

Dec 2025: 0 mentionsJan 2026: 0 mentionsFeb 2026: 0 mentionsMar 2026: 0 mentionsApr 2026: 0 mentionsMay 2026: 0 mentionsJun 2026: 5 mentionsJul 2026: 1 mentionAug 2026: 0 mentionsSep 2026: 4 mentionsOct 2026: 51 mentionsNov 2026: 0 mentions
Dec 2025Jun 2026Nov 2026
Mentions a month of the BuilderIO/agent-native repository that our weekly crawl found. There is no adoption data for this file yet.

Who's talking about BuilderIO/agent-native

The biggest 5 of 61 mentions linking to the repository that our crawl found. Each line is their own headline.

See all 61

Talked about on GitHub and Bluesky (61 times).

On GitHub · 50

45 more on GitHub

On Bluesky · 11

6 more on Bluesky

Does it work?

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

Reviews can't be read right now.

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.