g-review

g-review is a skill for Claude Code from kgiwojno/g-sdd. It costs 70 tokens per session (1,646 once invoked), scanned A, original, MIT.

A structured review of code changes against the feature specification and the project’s development instructions. It produces a report that labels findings by severity.

In plain words
What is it for?
Use it as a review gate for planned feature work or bug fixes, comparing the code diff with the requested behavior and recording findings for follow-up.
Why use it?
It catches missing requirements, rule violations, and other problems before an implementation slice or bug fix is accepted.

Skill for Claude Code

Written for Claude Code: allowed-tools in frontmatter. Also seen: mentions subagents; mentions AGENTS.md.

Part of the g-sdd plugin — 11 skills, 4 agents, 11 hooks shipped together

Good fit Use it as a review gate for planned feature work or bug fixes, comparing the code diff with the requested behavior and recording findings for follow-up.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/kgiwojno/g-sdd/g-review
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.

Any agent
npx skills add kgiwojno/g-sdd --skill g-review
Clone the repo
git clone --depth 1 https://github.com/kgiwojno/g-sdd

Made for: Claude Code.

Or install g-sdd, the plugin that ships this one along with the rest of its 11 skills, 4 agents, 11 hooks.

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 g-review

README.md
[![agentmods](https://agentmods.dev/badge/skills/kgiwojno/g-sdd/g-review/github.svg)](https://agentmods.dev/skills/kgiwojno/g-sdd/g-review)
Your own site
<a href="https://agentmods.dev/skills/kgiwojno/g-sdd/g-review"><img src="https://agentmods.dev/badge/skills/kgiwojno/g-sdd/g-review/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for g-review

Your own site · 80×15
<a href="https://agentmods.dev/skills/kgiwojno/g-sdd/g-review"><img src="https://agentmods.dev/badge/skills/kgiwojno/g-sdd/g-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 70 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,646 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00070 $0.01646
Opus 5 $0.00035 $0.00823
Sonnet 5 $0.00014 $0.00329
Haiku 4.5 $0.00007 $0.00165

Measured 12d ago against content hash fd97da7ffce5, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

g-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 12d 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.

skills/g-review/SKILL.md · 180 lines

How it starts

The opening of the file, as written. The whole thing — 180 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Behavior

/g-review is a per-change quality gate. It is invoked internally by /g-implement and /g-fix — never directly by the user in v1 (the ad-hoc contract is deferred).

Contract

Inputs (five named arguments from the caller):

  1. spec — what the change is supposed to do. For /g-implement, the full task entry from TASKS.md. For /g-fix, the bug description + approved fix approach.
  2. agents_md_path — path to the project's AGENTS.md.
  3. context_md_path — path to the project's CONTEXT.md (always passed by callers; the reviewer reads it at session start and applies the three-state axis matrix per its agent file).
  4. diff_range — git revision range (e.g. {feature-name}...{feature-name}-slice-NN-{slug}).
  5. output_path — where to write the review report.

Optional:

  1. prior_review_path — when re-reviewing after a re-spawn, the existing review report path. The skill reads the prior "Current Review" into a "Review History" entry before overwriting with the new iteration.
  2. extra_axis — extra project-specific axes to weight beyond the five standard axes. /g-implement passes scope-drift; /g-fix passes regression-test-presence. (The two new project-wide axes — glossary-alignment and context-md-spec-pollution — are baseline, not extra_axis; the reviewer applies them conditionally per the CONTEXT.md three-state matrix in its agent file.)

Output: a single review report file written to output_path, formatted per docs/templates/review-report.md.

Execution

Spawn the reviewing subagent:

cd into the worktree before spawning. The subagent inherits the orchestrator's cwd at invocation. Run this literally:

cd "$worktree_root"

Task tool invocation:

subagent_type: g-reviewer
description: Review {type} [NN] iter [N]
prompt: <<EOF
<!-- g-sdd:work-unit-type={type} -->
<!-- g-sdd:work-unit-number=[NN] -->
<!-- g-sdd:slice-branch={slice-branch} -->
<!-- g-sdd:worktree-path=/abs/path/to/worktree -->
<!-- g-sdd:feature-name={feature-name} -->
<!-- g-sdd:iteration=[N] -->

Read the full file on GitHub · 180 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. 12d ago First seen · 180 lines · 70 tokens per session scan A fd97da7ffce5

Subscribe to this mod's changes

g-review is a skill published in the GitHub repository kgiwojno/g-sdd (6 stars, last pushed 2mo ago), licensed MIT. It adds 70 tokens to every session and 1,646 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.