review-diff

A command that examines the changes currently shown by Git, the version-control system used to track code changes. It reports likely bugs and security problems, then recommends approval or more changes.

In plain words
What is it for?
Use it before committing or opening a pull request. It reports high-confidence findings with file and line details, separates bugs from minor comments, and ends with a clear decision.
Why use it?
It gives a focused review of the code you changed without burying important problems among guesses or minor style comments.

Command for Claude Code

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.

agentmods
npx agentmods add commands/ruvnet/metaharness/review-diff
Clone the repo
git clone --depth 1 https://github.com/ruvnet/metaharness

Made for: Claude Code.

Per session 13 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 77 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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.

ModelPer sessionOnce invoked
Fable 5 $0.00013 $0.00077
Opus 5 $0.00006 $0.00039
Sonnet 5 $0.00003 $0.00015
Haiku 4.5 $0.00001 $0.00008

Measured 3d ago against content hash 2ab52f01487b, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

review-diff 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.

kimi-k3-harness/.claude/commands/review-diff.md · 11 lines

What it actually says

Review the current git diff.

  1. git diff to read the change.
  2. Report only high-confidence findings as file:line — issue — fix.
  3. Separate bugs from nits.
  4. End with APPROVE or REQUEST-CHANGES and a one-line reason.
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.

  1. 3d ago First seen · 11 lines · 13 tokens per session scan A 2ab52f01487b

Subscribe to this mod's changes

review-diff is a command published in the GitHub repository ruvnet/metaharness (624 stars, last pushed 2d ago), licensed MIT. It adds 13 tokens to every session and 77 once invoked, about $0.0001 per session on Opus 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 2026-08-30.

Related

Other commands, from other repositories

prepare-this-repo

Prepare the current repository to abcd's working conventions — audit it against the abcd record, then adopt the three-tier .abcd/ layout, an AGENTS.md conventions section, and the commit gates. Interim bridge until abcd manages repos directly. Use when the user asks to prepare, onboard, scaffold, or bootstrap a repo…

intentdriven/abcd · 93 tokens

intent

Press-release intent lifecycle — status, quoted-text create, the implement-readiness gate, and the human planning interview that turns a draft into a planned, specced intent.

intentdriven/abcd · 37 tokens

ideate

Put a big, unproven idea through the three-leg admission gauntlet — primary-source research, a grill against the existing record, and an independent adversarial review — then record the verdict, whether the idea survives or dies.

intentdriven/abcd · 50 tokens

history

Manage the native session-transcript store for this repo by invoking the abcd binary. list, show and staged are read-only; capture and drain are the redacting write paths. The store is keyed on the repo's root-commit SHA and every stored transcript is redacted on write.

intentdriven/abcd · 60 tokens

memory

Query and curate the per-project memory substrate at .abcd/memory/ by invoking the abcd binary. Bare invocation is a read-only status render; ingest/ask/lint curate, synthesise, and health-check the store.

intentdriven/abcd · 48 tokens

nio

Nio — scan code, evaluate an action, read the audit report, manage config.

core0-io/nio · 17 tokens