review-loop

review-loop is a skill for Claude Code from EternallLight/tgc-skills. It costs 57 tokens per session (1,392 once invoked), scanned A, original, MIT.

A repeated code-review process that reviews a branch or pull request, fixes valid findings, and reviews it again until it is clean.

In plain words
What is it for?
Running authorized review-and-fix cycles on a branch or pull request, including one commit and push of the fixes, without merging it.
Why use it?
It avoids stopping after one review when fixes can introduce new problems or when legitimate issues remain.

Skill for Claude Code

Written for Claude Code: argument-hint in frontmatter. Also seen: mentions subagents.

Part of the tgc-skills plugin — 9 skills shipped together

Good fit Running authorized review-and-fix cycles on a branch or pull request, including one commit and push of the fixes, without merging it.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/eternalllight/tgc-skills/review-loop
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 EternallLight/tgc-skills --skill review-loop
Clone the repo
git clone --depth 1 https://github.com/EternallLight/tgc-skills

Made for: Claude Code.

Or install tgc-skills, the plugin that ships this one along with the rest of its 9 skills.

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 review-loop

README.md
[![agentmods](https://agentmods.dev/badge/skills/eternalllight/tgc-skills/review-loop/github.svg)](https://agentmods.dev/skills/eternalllight/tgc-skills/review-loop)
Your own site
<a href="https://agentmods.dev/skills/eternalllight/tgc-skills/review-loop"><img src="https://agentmods.dev/badge/skills/eternalllight/tgc-skills/review-loop/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 review-loop

Your own site · 80×15
<a href="https://agentmods.dev/skills/eternalllight/tgc-skills/review-loop"><img src="https://agentmods.dev/badge/skills/eternalllight/tgc-skills/review-loop.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 57 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,392 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 $0.00057 $0.01392
Opus 5 $0.00028 $0.00696
Sonnet 5 $0.00011 $0.00278
Haiku 4.5 $0.00006 $0.00139

Measured 12d ago against content hash 34b4c587edd6, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

review-loop 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/review-loop/SKILL.md · 138 lines

How it starts

The opening of the file, as written. The whole thing — 138 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Review Loop — review → fix → re-review until clean

Proceed only when the surrounding user request explicitly asks for the loop (including direct /tgc-skills:review-loop invocation) or tgc-skills:ship/tgc-skills:checkpoint invokes it. It commits fixes and pushes once, so stop without that authorization.

Review the branch diff with independent agents, fix legitimate findings, and repeat until clean or a stop condition is reached. Never merge the PR.

Keep the loop out of the main context

Create STATUS=$(mktemp "${TMPDIR:-/tmp}/tgc-review-loop.XXXXXX") and tell the user that progress is visible there. Then launch one general-purpose subagent with the repo path, optional PR/branch argument, status path, and the complete workflow below. The main context manages only the dispatch and final report. Subagents may start in the background; explicitly wait for this one to finish before returning or letting a parent workflow report completion.

Subagent workflow

Setup and heartbeat

  1. Resolve the local nameWithOwner, current branch, and GitHub default branch. Stop if the current branch is the default branch. Require a supplied branch argument to equal the current branch. If a PR is supplied or detected, require its URL repository to match nameWithOwner and its headRefName to match the current branch. Stop on any mismatch. Use its baseRefName; without a PR, use the default branch.

  2. Require a clean working tree. Stop and report staged, unstaged, or untracked files rather than silently excluding them.

  3. Fetch the resolved base and review its diff consistently throughout the loop. Capture the base ref from command output into a shell variable and always reference it quoted — a valid Git ref can contain shell metacharacters ($(...), backticks), and command-substitution output is not re-evaluated, so this cannot inject commands:

    BASE_REF=$(gh pr view <pr> --json baseRefName --jq .baseRefName \
      || gh repo view --json defaultBranchRef --jq .defaultBranchRef.name)
    git fetch origin "$BASE_REF"
    LOOP_BASE=$(git rev-parse HEAD)
    git diff "origin/$BASE_REF...HEAD"
    
  4. Create/truncate the passed status file outside the repo:

Read the full file on GitHub · 138 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 · 138 lines · 57 tokens per session scan A 34b4c587edd6

Subscribe to this mod's changes

review-loop is a skill published in the GitHub repository EternallLight/tgc-skills (8 stars, last pushed 1mo ago), licensed MIT. It adds 57 tokens to every session and 1,392 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.