Borrowing it
Nothing to install: this file belongs to incubateur-ademe/benefriches. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/incubateur-ademe/benefriches/main/.claude/commands/adw-review.mdgit clone --depth 1 https://github.com/incubateur-ademe/benefrichesWrote 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.
[](https://agentmods.dev/commands/incubateur-ademe/benefriches/adw-review)<a href="https://agentmods.dev/commands/incubateur-ademe/benefriches/adw-review"><img src="https://agentmods.dev/badge/commands/incubateur-ademe/benefriches/adw-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.
<a href="https://agentmods.dev/commands/incubateur-ademe/benefriches/adw-review"><img src="https://agentmods.dev/badge/commands/incubateur-ademe/benefriches/adw-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00018 | $0.00502 |
| Opus 5.5 | $0.00007 | $0.00201 |
| Sonnet 5.5 | $0.00004 | $0.00100 |
| Haiku 4.5 | $0.00002 | $0.00050 |
Grade A, and why
adw-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 9d 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.
What it actually says
Review the commits on the current branch since $2 against the spec at $1, and write the review to review.md in the spec's directory. Edit nothing else: no file in the worktree, no git state (no commit, checkout, stash, reset). You may run read-only and test commands (git diff, git log, focused tests, the spec's Validation commands) to confirm a finding.
- Record
git status --porcelain. It must be identical when you finish; if a command you ran changed it, say so inreview.md. - Read the spec, then
git log --oneline $2..HEADandgit diff $2...HEAD. - Invoke the
code-reviewerskill with the Skill tool, scopegit diff $2...HEAD, and follow it. - Check the diff against the spec: the Approach and every item under Files are implemented, or the deviation is explained; the tests listed under Validation exist.
- Classify every finding, citing
file:line:- blocker: the spec is not met, or a Critical or High finding that this diff introduces. Blockers send the branch back to the builder, so each one says what to change.
- tech debt: a Medium finding, or a real problem outside the spec's scope or that predates this diff. Each one becomes a follow-up issue, so it must stand alone: what, where, why it matters.
- Drop Low findings.
- Write
review.md: the verdict, the blockers, the tech debt, then what you checked (commands run and their results).
Approve only when there are no blockers. End your final message with the verdict, alone in a json code block:
{
"approve": false,
"blockers": [
{
"title": "...",
"location": "path:line",
"detail": "what is wrong and what to change"
}
],
"techDebt": [
{
"title": "...",
"location": "path:line",
"detail": "what, where, why it matters"
}
]
}
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.
- 9d ago First seen · 40 lines · 18 tokens per session scan A e6011d92c410
adw-review is a command published in the GitHub repository incubateur-ademe/benefriches (45 stars, last pushed today), licensed MIT. It adds 18 tokens to every session and 502 once invoked, about $0.0001 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 2026-10-21.
Other commands, from other repositories
elegant-code-review
Review the current working diff against the elegant-code decision ladder and propose deletions, honoring the negligence floor.
prp-commit
Quick commit with natural language file targeting: describe what to commit in plain English.
git
The pre-finish status: branch, hygiene findings, message checks, workflow lint, template state.
gsd:pr-branch
Create a clean PR branch by filtering out .planning/ commits — ready for code review.
validate-pr-description
Use when validating a PR title and description for conventional commit format, issue linking keywords, and template compliance before submission.
karpathy-check
Revisa as alterações em stage (ou o último commit) com base nos 4 princípios de codificação do Karpathy. Verifica complexidade, ruído no diff, suposições ocultas e verificação de objetivos. Uso: /karpathy-check [--last-commit].