code-review-give

code-review-give is a command for coding agents from axiomantic/spellbook. It costs 17 tokens per session (2,437 once invoked), scanned A, original, MIT.

A code-review mode for examining someone else’s changes, such as a pull request or branch, through several analysis passes.

In plain words
What is it for?
Use it to review pull requests, repository branches, or review URLs and produce file-and-line recommendations grouped by honest severity.
Why use it?
It provides evidence for each finding, checks every changed file, considers the project context, and reduces unsupported or duplicated feedback.

Command

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/axiomantic/spellbook/code-review-give
Clone the repo
git clone --depth 1 https://github.com/axiomantic/spellbook

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.

agentmods badge for code-review-give

README.md
[![agentmods](https://agentmods.dev/badge/commands/axiomantic/spellbook/code-review-give.svg)](https://agentmods.dev/commands/axiomantic/spellbook/code-review-give)
Your own site
<a href="https://agentmods.dev/commands/axiomantic/spellbook/code-review-give"><img src="https://agentmods.dev/badge/commands/axiomantic/spellbook/code-review-give.svg" alt="Measured on agentmods" height="20"></a>
Per session 17 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,437 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 $0.00017 $0.02437
Opus 5 $0.00009 $0.01218
Sonnet 5 $0.00003 $0.00487
Haiku 4.5 $0.00002 $0.00244

Measured 4d ago against content hash d9248504f559, 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-give 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 4d 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.

commands/code-review-give.md · 217 lines

How it starts

The opening of the file, as written. The whole thing — 217 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Code Review: Give Mode (--give <target>)

Invariant Principles

  1. Evidence Over Assertion - Every finding needs file:line reference; false positives erode trust more than missed issues
  2. Severity Honesty - CRITICAL=security/data loss/outage; HIGH=bugs and broken correctness; MEDIUM=quality concern or technical debt; LOW=minor improvement; NIT=purely stylistic; QUESTION=needs contributor input before judgment; PRAISE=noteworthy positive
  3. Context Awareness - Severity scales with risk surface: a timing issue in financial code is CRITICAL; the same code in a demo script is LOW
  4. Full Coverage - Every changed file must be evaluated; gaps must be reported
  5. Prior Context - Existing review threads inform the current review; do not duplicate or contradict unresolved feedback without justification

Target Formats

Target formats: 123 (PR#), owner/repo#123, URL, branch-name

Step 0: Load Project Conventions

When reviewing a PR (target is a PR number, URL, or fetched diff), the diff is the authoritative code. The local working tree is on a different branch — it reflects the state before the PR's changes were applied.

NEVER read local files that appear in the PR's changed file set. The local version is old code. Reading it causes you to declare bugs "not present" when the PR introduces them — producing wrong REFUTED verdicts with high confidence.

Local files are safe to read ONLY when:

  1. The file is not in the PR's changed file list, AND
  2. You are reading it for convention context only (not to verify PR behavior)

When in doubt, treat a file as changed and use the diff.

Before reviewing any code, load project context:

  1. If invoked within a develop run and design_context.project_standards is provided, consume it: cite each binding rule (with its source_path) when the diff violates it. This is the preferred source when available.
  2. Read the root AGENTS.md, plus every subdirectory AGENTS.md covering a changed path — a module's own conventions override the root's
  3. Read CLAUDE.md and/or .claude/CLAUDE.md if present in repo root
  4. Read docs/coding-standards.md and docs/ai/testing-instructions.md if present
  5. Read pyproject.toml, setup.cfg, .eslintrc, biome.json, ruff.toml, or equivalent style/lint config
  6. Read CONTRIBUTING.md or any style guide the repo references
  7. Check for docs/code-review-instructions.md or .github/code-review-instructions.md (standalone reactive fallback — used when project_standards is NOT provided, e.g. invoked outside a develop run)
  8. Sample 1-2 files adjacent to changed files to discover actual naming, style, and structural conventions — only files NOT in the PR's changed file set

Read the full file on GitHub · 217 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. 4d ago First seen · 217 lines · 17 tokens per session scan A d9248504f559

Subscribe to this mod's changes

code-review-give is a command published in the GitHub repository axiomantic/spellbook (10 stars, last pushed today), licensed MIT. It adds 17 tokens to every session and 2,437 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 2026-08-31.