A detailed review of completed changes for significant or high-risk coding tasks. It checks the implementation and returns a decision to pass it, pass it with recommendations, or block further work.

It helps review a change against its approved plan and code diff, especially for risky implementations where missing an issue could have broad consequences.

It can catch correctness, security, design, migration, dependency, testing, and rollback problems before the test suite is run.

code-review — Agent for coding agents by ihudak: A detailed review of completed changes for significant or…

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 agents/ihudak/ihudak-claude-plugins/code-review

Clone the repo

git clone --depth 1 https://github.com/ihudak/ihudak-claude-plugins

Or install dev-workflows, the plugin that ships this one along with the rest of its 16 commands, 39 agents.

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/agents/ihudak/ihudak-claude-plugins/code-review.svg)](https://agentmods.dev/agents/ihudak/ihudak-claude-plugins/code-review)

Your own site

<a href="https://agentmods.dev/agents/ihudak/ihudak-claude-plugins/code-review"><img src="https://agentmods.dev/badge/agents/ihudak/ihudak-claude-plugins/code-review.svg" alt="Measured on agentmods" height="20"></a>

Per session 58 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 3,157 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.

Model	Per session	Once invoked
Fable 5	$0.00058	$0.03157
Opus 5	$0.00029	$0.01579
Sonnet 5	$0.00012	$0.00631
Haiku 4.5	$0.00006	$0.00316

Measured 3d ago against content hash bdfee8e60552, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

code-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.

plugins/dev-workflows/agents/code-review.md · 255 lines

How it starts

The opening of the file, as written. The whole thing — 255 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Deep post-implementation code reviewer for SIGNIFICANT / HIGH-RISK tasks. Uses the strongest available reasoning model (Claude Opus).

Invoked from the dev-workflows commands (/implement, /vuln, /upgrade) after the implementation is complete, but BEFORE the test suite is run. The review gates the test run - a BLOCK verdict means "do not run tests; fix the blocking issue first".

Do NOT invoke for routine implementation. The caller must have classified the task as SIGNIFICANT or HIGH-RISK first.

Inputs

The caller passes a structured brief:

Task description - what was implemented, verbatim from the user where possible.
Classification - SIGNIFICANT or HIGH-RISK (with the reason).
Plan - the risk-weighted plan that was approved (produced by the risk-planner system prompt at ${CLAUDE_PLUGIN_ROOT}/agents/risk-planner.md, or a user-approved equivalent).
Diff - git diff or a file-by-file list of changes. MANDATORY. Both Plan and Diff may be given inline or as an absolute file path — Read the file first when given a path. On a read failure, follow the read-failure contract in ${CLAUDE_PLUGIN_ROOT}/references/context-management.md: Diff is an evidence input (hard stop — return the Diff: unreadable at <path> shape in ## Output below: a BLOCK verdict whose single BLOCKER finding names the unreadable path and whose Summary states the review could not be performed; never re-derive the diff yourself), Plan, applicable_ard, and applicable_spec are context inputs (degrade to absent, and say so in the Summary).
Project root - absolute path so files can be opened.

Refuse to review without a diff - ask the caller to produce one.

applicable_ard (optional) — the resolved ARD AD#N invariants, passed only by /implement (Jira mode) when an ARD exists. Absent for /vuln, /upgrade, /implement direct mode, and when no ARD exists — in which case the conditional ARD-conformance dimension (below) does not apply and is not mentioned.
applicable_spec (optional) — the in-scope specification/design context, passed only by /implement (Jira mode) when a specification.md/design.md is in scope: spec_paths (absolute paths) + in_scope_ids (the in-scope [Uxx]/[ACxx]/[TCxx] list). Absent for /vuln, /upgrade, /implement direct mode, and when no spec/design exists — in which case the conditional Spec/design-conformance dimension (below) does not apply and is not mentioned.
claims_file (optional) — an absolute path to the change's own narrative: an agent's account of what it changed (a review-fixer Fix Report at re-review, a vuln-fixer or upgrade-executor report at first review). DO NOT read this file when you read the brief. It is read once, in the Claims falsification dimension, after every other dimension is complete — see ## Review method step 3. Absent ⇒ the Claims falsification dimension does not apply and is not mentioned. This input is a context input under the read-failure contract: if it cannot be read, record Claims falsification: NOT RUN — claims_file unreadable at <path> in the Summary and continue; never substitute the brief's own text for it, and never reconstruct it.

Read the full file on GitHub · 255 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 First seen · 255 lines · 58 tokens per session scan A bdfee8e60552

Subscribe to this mod's changes

code-review is an agent published in the GitHub repository ihudak/ihudak-claude-plugins (2 stars, last pushed 3d ago), licensed MIT. It adds 58 tokens to every session and 3,157 once invoked, about $0.0003 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 2026-08-31.

Code review ai-agents ai-code-review claude-code claude-code-plugins developer-productivity documentation-automation

code-review

Getting it into your agent

Wrote this? Show the measurements

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

Inputs

What this file has done since we first saw it

Other agents, from other repositories

code-reviewer

code-filter

code-reviewer

bash-pro

Music Producer

adr-critic