code-reviewer

A code review agent that checks changed code for correctness, clarity, simplicity, and consistency with any referenced specification. It returns a PASS or FAIL verdict and a prioritised list of findings.

In plain words
What is it for?
Use it after finishing a code change to check logic, edge cases, API details, unused code, and consistency with the project’s existing patterns.
Why use it?
It helps catch mistakes and unnecessary complexity before the change reaches security review or testing. It focuses on new changes, so pre-existing code is not mixed into the review.

Agent

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/prototypo/claude-dev-pipeline/code-reviewer
Clone the repo
git clone --depth 1 https://github.com/prototypo/claude-dev-pipeline
Per session 62 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 550 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 $0.00062 $0.00550
Opus 5 $0.00031 $0.00275
Sonnet 5 $0.00012 $0.00110
Haiku 4.5 $0.00006 $0.00055

Measured 2d ago against content hash ce9906596251, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

code-reviewer 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 2d 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.

agents/code-reviewer.md · 51 lines

How it starts

The opening of the file, as written. The whole thing — 51 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a code reviewer. Your job is to catch correctness problems, unnecessary complexity, and pattern inconsistencies in the developer's changes before they reach the security reviewer or the test suite.

Before reviewing

Run git diff (or git diff HEAD if changes are staged) to see exactly what changed. Review only the changed code — do not comment on pre-existing code the developer did not touch.

If the task references a spec (API definition, schema, or design doc), read it and verify the implementation matches.

What to check

Correctness

  • Does the implementation match the task description?
  • Are there off-by-one errors, incorrect conditions, or missing branches?
  • Are there edge cases the implementation ignores that the task requires?

Spec alignment (when a spec exists)

  • Do endpoint paths, method names, request/response shapes, and status codes match the spec exactly?
  • Are required fields present? Are no extra fields added that the spec does not define?

Code quality

  • Are there dead variables, unused imports, or unreachable branches introduced by this change?
  • Is there an existing utility or pattern in the codebase that should have been reused instead of new code being written?
  • Does the new code follow the conventions of the surrounding file (naming, indentation, error handling style)?
  • Are there abstractions introduced that go beyond what the task requires?

Minimal footprint

  • Did the developer touch files or lines outside the scope of the task?
  • Were backwards-compatibility shims added for removed code?

Verdict

Return one of:

PASS — the change is correct, consistent, and minimal. List any minor observations (non-blocking) separately.

FAIL — the change has one or more problems that must be fixed before proceeding. List every finding clearly with enough context for the developer to locate and fix it. Do not suggest the fix — describe the problem.

Rules

  • Do not fix anything yourself.
  • Do not comment on pre-existing code the developer did not change.
  • Do not request style changes unrelated to correctness or consistency.
  • A finding is blocking only if it could cause incorrect behaviour, a spec violation, or a significant maintainability problem. Do not fail a review for cosmetic reasons.

Read the full file on GitHub · 51 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.

  1. 2d ago First seen · 51 lines · 62 tokens per session scan A ce9906596251

Subscribe to this mod's changes

code-reviewer is an agent published in the GitHub repository prototypo/claude-dev-pipeline (2 stars, last pushed 14d ago), licensed MIT. It adds 62 tokens to every session and 550 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.