agentleFS
DnsQueenDnsQueen/orcinus-skills/skills/finalize/SKILL.md

finalize

Skill1 starsChanged 5 days agoMITView on GitHub

Prepare a release, cut it, and hand it over — docs, release notes, CI gates, rollback plan, and monitoring — without publishing anything the user has not approved. Use when the work claims to be complete and ready to publish. Only invoke via the orcinus router.

22 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use simplify, debug and 2 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.
  • Hugging FaceSkill

    simplify

    Refactor code after tests pass. The "Refactor" phase of Red-Green-Refactor.

    Also in 3 projects that use this

  • AnthropicSkill

    debug

    Structured debugging session — reproduce, isolate, diagnose, and fix. Trigger with an error message or stack trace, "this works in staging but not prod", "something broke after the deploy", or when behavior diverges from expected and the cause isn't obvious.

    Also in 3 projects that use this

  • Microsoft AzureSkill

    verify

    End-to-end app testing pipeline for browser-based web apps on cloud browsers via Azure Playwright Workspaces: verify the live app, author tests for working flows, heal failing tests, then re-run with video and produce one consolidated report. Three phases (app verification, test authoring, test healing) plus a scoped final sanity run; each phase emits a sub-report. Use when asked to "verify my app", "make sure my app works and has good test coverage", "do a full QA pass", "add tests where I'm missing coverage", "fix failing tests and add new ones", "smoke test before deploy", "QA my deployment", or any combined verify-and-test ask. Requires Azure CLI (`az login`) and a Playwright Workspace resource ID — bootstraps configs on first invocation. Browser-based web apps with Playwright only — not for unit/API tests, native mobile/desktop, load/perf, or full accessibility audits.

    Also in 3 projects that use this

  • SentrySkill

    commit

    Create commit messages following Sentry conventions. Use when committing code changes, writing commit messages, or formatting git history. Follows conventional commits with Sentry-specific issue references.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 6 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 22 edited versions we read keep these.

  • 6

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  • could be automatic: a setting could enforce it
  1. Distinctive:

    No publish without explicit human confirmation.

    line 14
  2. Distinctive:

    No claim without evidence.

    line 15
  3. Distinctive:

    Use ready tools as-is. Record each tool with its version, or "not run" with the reason.

    line 16
  4. Very specific:

    Use the installed handoff skill as-is.

    line 27
  5. Distinctive:

    Write down, with evidence for each item:

    line 39
  6. Distinctive:

    No secrets in the release, no failing checks.

    line 45
  7. Distinctive:

    Stop here. Present the release report (below) and ask the user to confirm it in writing.

    line 53

“Common” is measured: kept by at least 6 in 10 of the 22 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

FINALIZE — ready to publish, and provably so

Prepare the release and hand it over. Do not publish, merge, or tag anything until the user confirms in writing. Use the ready tools as-is — do not rewrite them, and never claim one ran when it did not.

Rules

  • No publish without explicit human confirmation. No auto-merge, no auto-publish, no auto-tag.
  • No claim without evidence. Every item in the release report points to a command, an output, or a file.
  • Use ready tools as-is. Record each tool with its version, or "not run" with the reason.
  • A check that did not run is never reported as passed.

Steps, in order

1. Prerequisites

Confirm orcinus-verify gave PASS and security-review gave PASS (or explicit exceptions) on the change. If either is FAIL or INCOMPLETE, stop and fix that first. Finalize does not override verification or security.

2. Docs and handoff

Use the installed handoff skill as-is. Write the handoff: what changed, how to run the project, what is next, what is blocked, and what must not be forgotten. Do not rewrite handoff; use it.

3. Release notes — use release-please as-is

Use release-please (googleapis/release-please, Apache-2.0, CLI + GitHub Action) as-is to generate release notes from conventional commits. Do not hand-write notes when release-please can generate them. If release-please is not set up in the project, record "not run — release-please not set up" and write the notes by hand from the git log, marking them as hand-written. Never claim release-please ran when it did not. Note: release-please does not publish to package registries; registry publishing is a separate step, not covered here.

4. CI gates

Re-run the four commands (build, lint, typecheck, test) and record their outputs. Use Waza (skills/waza, microsoft/waza, MIT, CLI) as-is for graded eval scenarios when the project has them (the repo ships a runnable example, waza-orcinus-verify, for a full-repo install). Record each result with its tool and version, or "not run" with the reason.

5. Release checklist

Write down, with evidence for each item:

  • Version bumped (name the version).
  • Release notes (generated or hand-written, and which).
  • Docs updated.
  • Rollback plan: how to revert, who does it, and how long it takes. A release without a rollback plan is not ready.
  • No secrets in the release, no failing checks.

6. Monitoring after publish — use the ready skills as-is

Two skills cover monitoring, and they are complements, not substitutes. Use monitoring (injected verbatim, MIT — see its ATTRIBUTION.md) for the outside view: uptime probes on the critical journey, /livez split from /readyz, SLO burn-rate alerts with two windows, on-call escalation with a runbook per alert. Use monitoring-observability (injected verbatim, MIT, OrchestKit — see its ATTRIBUTION.md) for the telemetry side: Prometheus, Grafana, Langfuse, or LLM tracing. When the project uses neither (for example, a library with no runtime), say so plainly and record that the project has no runtime to monitor. Do not claim monitoring exists when it does not.

7. Gate: Code → Done / Publish

Stop here. Present the release report (below) and ask the user to confirm it in writing. Do not publish, merge, or tag until they confirm. If they change anything, update the report and confirm again.

8. After publish

New issues go through bug triage. The ready skill for that is issue-triage (MIT, copied verbatim into this repo — see its ATTRIBUTION.md). Record new issues in the project's tracker with the same evidence standard: what broke, how to reproduce it, and what the last known-good version was.

Release report

Return the report in this form, no more, no less:

  • Version.
  • Release notes (generated by release-please, or hand-written from the git log — and which).
  • Verification: orcinus-verify verdict, security-review verdict, four commands with outputs. A security-review verdict of PASS WITH LIMITATIONS is not a PASS: list each limitation, say whether it blocks the release, and record the user's decision on each one before confirming.
  • CI gates (Waza, if run).
  • Rollback plan.
  • Monitoring plan (monitoring for the outside view, monitoring-observability for telemetry, minimal signals, or no runtime — and which).
  • Handoff (what changed, how to run, what is next, what is blocked).
  • What was explicitly not published, and why.

Sources

This skill is an orchestrator. It uses ready tools as-is: release-please (googleapis/release-please, Apache-2.0, licence verified from the primary LICENSE file); Waza (microsoft/waza, MIT); handoff (installed, MIT); monitoring-observability (yonatangross/orchestkit, MIT, injected verbatim — see its ATTRIBUTION.md). No text was copied from release-please, Waza, or handoff. Issue-triage is copied verbatim (MIT — see its ATTRIBUTION.md).

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.