final-review
Skill2 starsChanged 5 days agoMITView on GitHub
Use for a final, release-gate review of a repository, project, app, tool, kit, documentation set, workflow, release candidate, or any deliverable before it is considered done or shipped. Checks structure, completeness, coherence, documentation, commands, tests, security, secrets, installation, usability, and release readiness, then returns a fixed-format report with a clear verdict - PASS, PASS WITH RISKS, or FAIL. Trigger on requests like "final review", "pre-release check", "is this ready to ship", or "review before delivery".
8 other projects use a version of this.
Adopted by
Most recently taken up by
- markmhendrickson/neotomatheir own version, first seen Nov 15, 2026
- normalled/apijacktheir own version, first seen Nov 8, 2026
- MrD-Lt/codex-final-reviewtheir own version, first seen Nov 7, 2026
- DanMcInerney/architect-looptheir own version, first seen Oct 8, 2026
- andeeplus/aplustheir own version, first seen Oct 4, 2026
- jsuvic/agent-profiletheir own version, first seen Aug 5, 2026
Other versions8 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 8
- DanMcInerney/architect-loop★ 626same name
- markmhendrickson/neotoma★ 38same name
- aromanarguello/roman-skills★ 18same name
- normalled/apijack★ 7same name
- limpidwei/final-exam-review★ 5same name
- jsuvic/agent-profile★ 3same name
- andeeplus/aplus★ 2same name
- MrD-Lt/codex-final-review★ 2same 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.
The file
We split this skill into the rules it gives an agent
10 rules. 0 are common, 10 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 7 edited versions we read keep these.
10
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 17
Follow these steps in order.
- Distinctive: line 20
Structure and completeness: Check the file tree: expected files present (README, LICENSE, manifests, entry points, configs), nothing …
- Distinctive: line 21
Coherence: Verify that code, configuration, and documentation agree with each other: names, versions, paths, commands, and described …
- Distinctive: line 22
Documented commands: Every command in the README and docs (install, build, run, test) must work.
- Distinctive: line 23
Tests, build, lint: Run the available test, build, lint, and validation steps and record the exact results.
- Distinctive: line 26
Release readiness: Check versioning, changelog, licensing, and any residual TODO/FIXME or open question that blocks release.
- Distinctive: line 30
Always return the report in this exact structure:
- Distinctive: line 54
Do not modify files unless explicitly requested.
- Distinctive: line 56
Prefer a short, hard finding over a long, soft one.
- Distinctive: line 57
Every claim in the report must be verifiable from the deliverable itself.
“Common” is measured: kept by at least 6 in 10 of the 7 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
Final Review
Review the deliverable like a strict technical supervisor signing off a release. Be practical and specific. Do not philosophize. Do not modify any files unless explicitly asked to fix something.
Scope
Works on any deliverable: code repositories, applications, CLI tools, libraries, kits and templates, documentation sets, automation workflows, release candidates, or individual changes (diffs and pull requests).
Procedure
Follow these steps in order. Skip a step only if it does not apply, and say so in the report.
- Define the deliverable. Identify what is being reviewed (repo, diff, app, docs, kit) and what "done" means for it. If reviewing a change, inspect the git diff and the list of modified, added, and removed files.
- Structure and completeness. Check the file tree: expected files present (README, LICENSE, manifests, entry points, configs), nothing missing, no leftovers (temp files, build artifacts, dead code, unused dependencies, commented-out blocks).
- Coherence. Verify that code, configuration, and documentation agree with each other: names, versions, paths, commands, and described behavior must match reality.
- Documented commands. Every command in the README and docs (install, build, run, test) must work. Run them when possible; otherwise verify them statically against the project.
- Tests, build, lint. Run the available test, build, lint, and validation steps and record the exact results. If none exist, flag their absence as a finding.
- Security and secrets. Scan for hardcoded credentials, API keys, tokens, private URLs and paths, personal data, and risky patterns (unpinned remote scripts, overly broad permissions, injection-prone constructs). Nothing secret may ship.
- Installation and usability. Confirm a newcomer can install and use the deliverable by following the documentation alone. Verify entry points and the quick-start path.
- Release readiness. Check versioning, changelog, licensing, and any residual TODO/FIXME or open question that blocks release.
Output format
Always return the report in this exact structure:
## Final Review
1. What was reviewed / what changed
2. Files and structure
3. Matches the request/spec: yes/no + gaps
4. Checks run (tests, build, lint, commands) + results
5. Security & secrets findings
6. Risks left
7. Fixes required before release
Verdict: PASS | PASS WITH RISKS | FAILVerdicts
- PASS — ready to ship. No blocking issues, no unmitigated risks.
- PASS WITH RISKS — shippable, but named risks remain. List every risk with its concrete impact; the owner decides whether to accept them.
- FAIL — not shippable. At least one blocking problem (broken command, missing critical file, exposed secret, failing test, docs that contradict the code). List exactly what must be fixed, in priority order.
Rules
- Do not modify files unless explicitly requested. Reviewing and fixing are separate jobs.
- Judge only against evidence: files, diffs, command output. No assumptions, no benefit of the doubt.
- Prefer a short, hard finding over a long, soft one.
- Every claim in the report must be verifiable from the deliverable itself.
- Missing information (no tests, no docs) is a finding to report, never a reason to stop the review.
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.

