Skip to content
Skills Library
Programming · Review & audit

review-diff

/review-diff

Two-axis review of the changes since a fixed point (commit, branch, tag, merge-base). Standards checks the diff against the repo's documented conventions plus a code-smell baseline; Spec checks it against the feature specs it claims to implement. Both run as parallel subagents and are reported side by side. Use when the user wants to review a branch, a PR, work in progress, or "review since X", and when /implement closes a phase.

Repository
rubstack ↗
Licence
MIT
Last repository change
08.11.2026

Assessment and security

95/ 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.
Content30 / 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
2
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 08.11.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Two-axis review of the changes since a fixed point (commit, branch, tag, merge-base). Standards checks the diff against the repo's documented conventions plus a code-smell baseline; Spec checks it against the feature specs it claims to implement. Both run as parallel subagents and are reported side by side. Use when the user wants to review a branch, a PR, work in progress, or "review since X", and when /implement closes a phase.

Triggers from the description

review since X
Original author description

Two-axis review of the changes since a fixed point (commit, branch, tag, merge-base). Standards checks the diff against the repo's documented conventions plus a code-smell baseline; Spec checks it against the feature specs it claims to implement. Both run as parallel subagents and are reported side by side. Use when the user wants to review a branch, a PR, work in progress, or "review since X", and when /implement closes a phase.

How it works

  1. Step 1: Pin the fixed point
  2. Step 2: Find the sources
  3. Step 3: Spawn both reviewers in parallel
  4. Step 4: Report
Original SKILL.md headings · 33 lines, 410 words
SKILL.md preview
---
name: review-diff
description: Two-axis review of the changes since a fixed point (commit, branch, tag, merge-base). Standards checks the diff against the repo's documented conventions plus a code-smell baseline; Spec checks it against the feature specs it claims to implement. Both run as parallel subagents and are reported side by side. Use when the user wants to review a branch, a PR, work in progress, or "review since X", and when /implement closes a phase.
---

# Review diff

Two separate questions about one diff:

- **Standards**: does the code follow this repo's documented conventions?
- **Spec**: does the code do what its feature spec asks, no more and no less?

A change can pass one and fail the other, so they run in separate subagents and are never merged into one ranked list. This is a read-only review: it reports, it doesn't fix. For behavioral bugs, use [/interrogate](../interrogate/SKILL.md).

## Step 1: Pin the fixed point

Use what the user named. Otherwise use the merge base with the default branch. Resolve it (`git rev-parse <ref>`) and confirm `git diff <ref>...HEAD` is non-empty before spawning anything; a bad ref or empty diff stops here. Note the diff command and `git log <ref>..HEAD --oneline`.

## Step 2: Find the sources

**Spec**, in order: a path the user passed; the sections of `.rubstack/engineering/features.md` that match the commit messages, branch name or touched files; issue references in commit messages (`#123`). None found → the Spec axis reports "no spec available".

**Standards**: `AGENTS.md`, `CONTRIBUTING.md`, `CODING_STANDARDS.md`, `.rubstack/adr/`, and any other file that says how code here should be written. Also use the exact terms in `.rubstack/product/CONTEXT.md` if it exists. On top, always apply the smell baseline in [REFERENCE.md](REFERENCE.md). Repo standards override the baseline, and anything a linter or typechecker already enforces is skipped.

## Step 3: Spawn both reviewers in parallel

One message, two Agent calls (`subagent_type: general-purpose`), each with a fresh context and file paths rather than pasted files. Each gets the diff command, the commit list and its sources, plus its brief from [REFERENCE.md](REFERENCE.md). Tell both: read-only, don't edit files.


## Step 4: Report

Present the two reports under `## Standards` and `## Spec`, lightly cleaned, not merged or re-ranked. End with one line: findings per axis and the worst one within each axis. Don't name a single worst finding across both axes.

Spec gaps on a feature marked `done` are the most important output: they mean `/qa` will test something that isn't there. When a calling skill says what to do with them (`/implement` marks the feature `blocked`), follow the caller. Run on its own, route them: small and contained → `/task`; unclear spec → `/evolve` or `/dev-plan`.

Example: `/review-diff main` on a branch that adds CSV export ends with "Standards: 2 findings, worst: possible Duplicated Code in `export.ts` and `report.ts`. Spec: 1 finding, worst: criterion 'Given an empty table, When exporting, Then a header-only file downloads' isn't implemented."
Files in this package (2)
  • REFERENCE.md 2 KB
  • SKILL.md 3 KB

Versions

RepositoryScoreLicence
This page · RubenGlez/rubstack95MIT
se-uhd/ai-slop-skill100CC-BY-4.0
zanderadam/engy100MIT
jisuzzzz/haechi100MIT
AmirAzy/HoNspot100MIT

Database snapshot: 13.11.2026 · package version 2aa9e7f72f67 (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

  • 88code-review
    Two-axis review of a git diff since a fixed point (commit, branch, tag, merge-base). Standards axis checks repo coding standards + Fowler smell baseline; Spec axis checks diff against originating issue/spec. Runs both axes as parallel sub-agents, aggregates side-by-side. Use for…
  • 90two-axis-review
    Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes: Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating ticket or spec asked for?). Runs both reviews in parallel…
  • 90review-code
    Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes — Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating spec/PRD asked for?). Use when the user wants to review a…
  • 90review-changes
    Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes: Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the plan, ticket or spec asked for?). Runs both reviews in parallel sub-agents…
  • 95standards-spec-review
    Two-axis review of the changes since a fixed point (commit, branch, tag, merge-base, or the working tree) on Standards (does the code follow this repository's documented coding standards, design documents, and the smell baseline?) and Spec (does the code do what the originating…
  • 90code-review
    Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes — Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/spec asked for?). Runs both reviews in parallel…

Feedback and reports

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