Skip to content
Skills Library
Programming · Review & audit

code-review

/code-review

Practice skill: review a diff against its Spec. Use in an aidd repository when reading a PR/MR as a reviewer, or someone asks to review a diff or whether it is ready to merge.

Repository
aidd ↗
Licence
Apache-2.0
Last repository change
12.11.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
4
Scripts
0
HTML comments
2
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 12.11.2026, without references to retired models or APIs.

Licence

Apache-2.0, otwarta: można używać i zmieniać, z zachowaniem licencji i informacji o zmianach.

What to check before installing
  • SKILL.md zawiera komentarze HTML (2). Nie widać ich w podglądzie Markdown, a model je czyta.

What it does and when to use it

Practice skill: review a diff against its Spec. Use in an aidd repository when reading a PR/MR as a reviewer, or someone asks to review a diff or whether it is ready to merge.

Original author description

Practice skill: review a diff against its Spec. Use in an aidd repository when reading a PR/MR as a reviewer, or someone asks to review a diff or whether it is ready to merge.

How it works

  1. Standards
  2. Review
  3. Test commands
  4. Stop if
  5. Hand off
Original SKILL.md headings · 52 lines, 587 words
SKILL.md preview
---
name: code-review
description: "Practice skill: review a diff against its Spec. Use in an aidd repository when reading a PR/MR as a reviewer, or someone asks to review a diff or whether it is ready to merge."
metadata:
  checked: "2026-10-22"
---

# Code review

**Use it when** you read a pull request as the reviewer, or as second reader for the security engineer or the product designer.

**Do not use when** you are writing the change, not reading someone else's. A self-review before handoff is Implement's own verification.

## Standards

<!-- aidd:gen practice:code-review -->
| Standard | Edition |
| --- | --- |
| Google Engineering Practices: the code review standard | — |
| Conventional Comments labels (emerging) | — |
| Diff discipline: every changed line traces to the Spec, and speculative, single-use, unrequested, or drive-by code is a finding | — |
| The OWASP Code Review Guide | — |
| AI-assisted coding: verify every generated line, API, and package before relying on it, and keep secrets and personal data out of prompts | — |

Each standard's checked date and source are in the aidd-team roles reference (`references/roles.md` in `aidd-team:aidd-team`), section `aidd-team:code-review`.
<!-- aidd:end -->

## Review

- Read the Spec, then the diff, then the checks on the head (`aidd test` evidence, or the platform's checks).
- Judge design, functionality, complexity, tests, naming, comments, style, and docs. Say what would improve code health, not what you would have written. "This name made me guess wrong" is a valid blocking comment (`aidd-team:naming`). A diff's shape and its comments follow `aidd-team:maintainable-code`.
- Spec conformance: every Acceptance line has a test that would catch a regression. Nothing is outside the scope. Rollback, observe, and decisions match what landed.
- Diff discipline: a speculative feature, a single-use abstraction, unrequested configurability, handling an impossible case, or a drive-by edit is a finding, and so is deleting pre-existing dead code unasked.
- Security trace: follow untrusted input to its sinks. Check authorization on every new path, secrets, injection, deserialization, and rate limits. Shape each security finding, its severity, and its outcome as [references/security-findings.md](references/security-findings.md) says. Escalate a major-route change to the security engineer.
- Ask which test compares a changed reader, parser, or decider's new answers with the old over a generated set, and which shows that the honest work a new control could block still passes.
- Run the passes in [references/passes.md](references/passes.md), over the artifact, a stated rule, untrusted text, AI-generated code, silent failures, test strength, and operability.

### Report

- Report as [references/review-output.md](references/review-output.md) says: both questions, and the verdict first. Rank findings Critical, High, Medium, Low, Nit. Give each its file, line, failure scenario, and suggested fix, quote the line it is about, and give a confidence from 1 to 10 (REVIEW.md sets both rules). When a finding feels vague or its severity unclear, compare it with the worked examples there. Label each with Conventional Comments (issue, suggestion, question, nitpick), `blocking` or `non-blocking`. Fold duplicates. Name disagreements.
- Never approve, merge, or edit. Run only commands that read.

## Test commands

The reviewer adds no test commands. It runs the touched tests and reads the rest in the evidence.

- From `aidd-team:test-design`: Mutation run.

## Stop if

- The diff is too large to review line by line: ask for a split rather than skim it.
- A finding needs a judgment only the person who decides Merge approval can make.
- You are about to edit the change, approve it, or merge it.

## Hand off

- Done: the ranked report. Next role: the author posts it and a person decides Merge approval. Findings needing code return to Implement. Needs: a judgment only that person can make. Stopped at: none, or the stop rule's code.
Files in this package (4)
  • SKILL.md 4 KB
  • references/passes.md 4 KB
  • references/review-output.md 6 KB
  • references/security-findings.md 4 KB

Versions

RepositoryScoreLicence
This page · thaithienvanid/aidd100Apache-2.0
microsoft/aspire100MIT
kdlbs/kandev100AGPL-3.0
dotnet/runtime100MIT
drosam/agents100MIT
ggml-org/llama.cpp100MIT
ShamelesAbyss/Nyx100Apache-2.0
dzhng/jevgrep100MIT
iuliandita/skills100MIT

Database snapshot: 13.11.2026 · package version b5c181ece612 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100aidd
    Use before any change in a repository with .aidd/config.yaml (or an AGENTS.md that names aidd): a bug fix, feature, refactor, commit, push, or PR/MR. Also when the user names aidd, a Spec, a route (minor, normal, major, emergency), or a risk rule.
  • 100review-github-pr
    Review a GitHub pull request by summarizing its diff and key design decisions. Use when the user wants to review a PR, understand changes in a branch, or get a code review summary. Trigger keywords - review PR, review pull request, summarize PR, summarize diff, code review,…
  • 90review
    Review a pull request or diff for this repository against otari's project standards. Use when asked to review a PR, inspect a diff, or prepare review feedback.
  • 90reviewing
    Reviews a diff against the project's standards and spec in priority order, ending in a ship or fix-first verdict. Use when the user asks to review code, a PR or a branch, "is this ready to merge?", or for UI code feedback. Not for debugging.
  • 100dev-review
    Use when a diff, branch or PR needs review before merge, on correctness, standards and the spec.
  • 100visual-diff-review
    Use when asked to review a diff and produce a 7-section visual page. Not for interactive review walks or PR-specific review: use show-review.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.