A code-review command that examines either your local Git changes or a specific pull request using a general or specialist review focus.

Use it to review current work, inspect a pull request, or check whether feedback was addressed in a follow-up review.

It helps find security, accessibility, dependency, performance, test, user-experience, or contributor-trust problems before changes are merged.

review — Command for Claude Code by lugassawan/swe-workbench

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Clone the repo

git clone --depth 1 https://github.com/lugassawan/swe-workbench

Made for: Claude Code.

Or install swe-workbench, the plugin that ships this one along with the rest of its 61 skills, 25 commands, 23 agents, 4 hooks.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/commands/lugassawan/swe-workbench/review/github.svg)](https://agentmods.dev/commands/lugassawan/swe-workbench/review)

Your own site

<a href="https://agentmods.dev/commands/lugassawan/swe-workbench/review"><img src="https://agentmods.dev/badge/commands/lugassawan/swe-workbench/review/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/commands/lugassawan/swe-workbench/review"><img src="https://agentmods.dev/badge/commands/lugassawan/swe-workbench/review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 72 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 4,762 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00072	$0.04762
Opus 5.5	$0.00029	$0.01905
Sonnet 5.5	$0.00014	$0.00952
Haiku 4.5	$0.00007	$0.00476

Measured 3d ago against content hash 8dec356d126c, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

review scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

commands/review.md · 146 lines

How it starts

The opening of the file, as written. The whole thing — 146 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Review code with senior-engineer depth. Two dimensions — fully orthogonal:

Auditor axis (--mode): which specialist reviews the diff (general / security / accessibility / dependency / performance / tests / contributor-trust). Auto-inferred from the diff when omitted.
Diff-source axis: local working-tree diff vs. PR diff. Determined by the remaining arguments after --mode is stripped.

Step 1 — Argument resolution

Parse $ARGUMENTS left-to-right:

If --approve-over-open-threads "<reason>" is present, strip it from $ARGUMENTS first — before any other flag parsing below, so it survives regardless of which mode is subsequently selected — and carry <reason> through as $APPROVE_OVER_OPEN_THREADS to whichever delegation this command makes to swe-workbench:workflow-pr-review: both the PR-mode MODE=auto delegation (PR mode below) and the Followup-mode MODE=followup delegation (Followup mode below), since the skill's Step 5.5 reads $APPROVE_OVER_OPEN_THREADS to pre-answer its own AskUserQuestion prompt. Absent this flag, $APPROVE_OVER_OPEN_THREADS stays empty and Step 5.5 prompts normally. This flag is meaningless outside general/--check-followup mode — a postable specialist mode or contributor-trust never reaches swe-workbench:workflow-pr-review, so it's silently unused there.

When this flag is safe to use: only after Step 5.5 (in skills/workflow-pr-review/SKILL.md) has actually run its verification pass and reported which threads remain open. It exists to answer — or pre-empt — the AskUserQuestion prompt Step 5.5 raises once genuinely-unresolved threads remain, for a thread the PR owner deliberately chose to leave open (e.g. a deferred follow-up). It is not a blanket way to force an approval without ever looking at what's still unresolved — supplying it on a first invocation just skips the prompt on faith that nothing in the still-open set actually needs a look.

0.5. If --check-followup <N> is present (where N is a PR number), strip it and enter Followup mode — see ## Followup mode below. All other flags and argument parsing (besides item 0 above, already stripped) are skipped.

If a --mode <value> flag is present, extract it and normalize the alias:

`--mode` value (and aliases)	Normalized mode	Delegates to
`general`	`general`	`swe-workbench:reviewer`
`security`, `sec`	`security`	`swe-workbench:security-auditor`
`accessibility`, `a11y`	`accessibility`	`swe-workbench:accessibility-auditor`
`dependency`, `deps`	`dependency`	`swe-workbench:dependency-auditor`
`performance`, `perf`	`performance`	`swe-workbench:performance-tuner`
`tests` (no short alias — keyword is already short)	`tests`	`swe-workbench:test-reviewer`
`contributor-trust`, `trust`	`contributor-trust`	`swe-workbench:contributor-auditor`
`ux` (no short alias)	`ux`	`swe-workbench:product-designer`

Strip --mode <value> from $ARGUMENTS. Store the normalized mode. If the value is unrecognized, print an error listing valid values and stop.

The remaining $ARGUMENTS (after stripping --mode) flow into diff-source detection:
- Matches [1-9][0-9]* (stripping a leading # if present) → PR mode with that number.
- Else, run gh pr view --json number,headRefName 2>/dev/null. If it succeeds (current branch has an open PR), print:
  
  "Detected PR #N on this branch — review it? Reply yes to enter PR mode, or local to review the local diff instead."
  
  Wait for the user's reply. yes → PR mode. local (or anything else) → local-diff mode.
- Else → local-diff mode.

Read the full file on GitHub · 146 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

3d ago Changed 8dec356d126c
27d ago Changed 6cade73d48f4
1mo ago Changed · +4 lines df62021003b9
1mo ago First seen · 142 lines · 72 tokens per session scan A ea5953da1fc0

Subscribe to this mod's changes

review is a command published in the GitHub repository lugassawan/swe-workbench (3 stars, last pushed 2d ago), licensed MIT. It adds 72 tokens to every session and 4,762 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Claude Code Code review Design and UX Go Testing claude-code-plugin clean-architecture design-patterns domain-driven-design

local-code-review-tests

Review uncommitted changes, build projects, and run unit tests.

devMappouras/local-code-review-claude-plugin · 13 tokens

octo-deliver

Delivery phase - Review, validate, and test with multi-AI quality assurance.

nyldn/claude-octopus · 15 tokens

pre-merge-check

Tu es un assistant de qualité code. Tu dois effectuer toutes les vérifications nécessaires AVANT de merger une branche, pour garantir la qualité et éviter les régressions.

TheBeardedBearSAS/claude-craft · 6 tokens

tdd

Test-driven development with tests that actually catch breaks: red before green, name the break each test catches, and the mutation check before done.

azrtydxb/procoder · 28 tokens

ia-test-browser

Run browser tests on pages affected by current PR or branch.

iliaal/whetstone · 15 tokens

ux-grill

Use when a running web app's UX has to be audited reproducibly rather than by feel — a deterministic mechanical pass over routes and viewports (axe, target size, horizontal overflow, page title, scripted journeys) followed by a screenshot-grounded interrogation of the operator, journey by journey, with two persona…

Kanevry/session-orchestrator · 121 tokens