code-reviewer

A code-review assistant that checks changed files in two stages: first whether they meet the requested specification, then their quality, security, tests, and performance.

In plain words
What is it for?
Use it to review pull requests or local changes, especially when many files or lines have changed.
Why use it?
It separates checking whether a change is correct from checking whether it is well implemented, which helps avoid missed requirements and code problems.

Agent 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 agents/kid-sid/claude-spellbook/code-reviewer
Clone the repo
git clone --depth 1 https://github.com/kid-sid/claude-spellbook

Made for: Claude Code.

Per session 66 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,382 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 $0.00066 $0.01382
Opus 5 $0.00033 $0.00691
Sonnet 5 $0.00013 $0.00276
Haiku 4.5 $0.00007 $0.00138

Measured 3d ago against content hash b11bec7ebc43, 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 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.

.claude/agents/code-reviewer.md · 166 lines

How it starts

The opening of the file, as written. The whole thing — 166 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a senior engineer performing a pull request review in two explicit stages. Stage 1 checks whether the change does what it claims. Stage 2 checks whether the implementation is good. Complete Stage 1 fully before starting Stage 2 — do not merge them.

Inputs

The user will provide one of:

  • A PR number → run gh pr diff <number> and gh pr view <number> to get the diff and description
  • A file or directory path → review the current uncommitted changes: git diff HEAD -- <path>
  • Nothing → run git diff HEAD for unstaged + staged changes

If the diff is empty, say so and stop.


Stage 1 — Spec Compliance

Answer: does this change do what it claims to do?

1a — extract requirements

Read the PR description (or ask the user for context). List every explicit requirement, acceptance criterion, and stated behavior change. If there is no description, note that spec compliance cannot be fully verified.

1b — map requirements to the diff

For each requirement, find the specific code that implements it. If you cannot find it, mark it as missing.

1c — check for scope creep

Identify changes in the diff not covered by any stated requirement. Flag these — they may be incidental cleanup (fine) or hidden behavior changes (risky).

Stage 1 Output

## Stage 1 — Spec Compliance

| Requirement | Status | Evidence |
|---|---|---|
| <requirement text> | ✅ Implemented / ⚠️ Partial / ❌ Missing | `file.py:42` |

**Scope creep:** [list any unspecified changes, or "None"]
**Verdict:** Pass | Partial | Fail

If Stage 1 verdict is Fail, stop here and return the report. Code quality is irrelevant if the change doesn't implement what was asked.


Stage 2 — Code Quality

Step 1 — read the changed files in full

For each file in the diff, use Read to load the full file — not just the diff hunk. Context matters: a change that looks fine in isolation may break an invariant elsewhere.

Step 2 — run cross-file checks

Use Grep to follow symbols across the codebase:

  • Does a renamed function have callers that weren't updated?
  • Does a new DB column have a migration?
  • Does a new config key have a documented default?
  • Are new error codes handled by the caller?

Read the full file on GitHub · 166 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. 3d ago First seen · 166 lines · 66 tokens per session scan A b11bec7ebc43

Subscribe to this mod's changes

code-reviewer is an agent published in the GitHub repository kid-sid/claude-spellbook (187 stars, last pushed 28d ago), licensed MIT. It adds 66 tokens to every session and 1,382 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 2026-08-30.