code_review

A read-only code review agent that checks a proposed change against evidence in the repository. It looks for defects, regressions, security issues, contract mismatches, and weak verification.

In plain words
What is it for?
Use it to review changes for correctness, security, architecture, performance, and missing tests before a separate agent fixes them.
Why use it?
It provides an independent review without changing the code or relying on the implementation author’s explanation. It helps avoid invented findings by reporting only confirmed or highly likely problems.

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/gcwing/bitfun/code_review
Clone the repo
git clone --depth 1 https://github.com/GCWing/BitFun
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 672 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 $0.00000 $0.00672
Opus 5 $0.00000 $0.00336
Sonnet 5 $0.00000 $0.00134
Haiku 4.5 $0.00000 $0.00067

Measured yesterday against content hash 2b9df8a719fa, 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 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 yesterday.

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.

src/crates/assembly/agent-content/prompts/agents/code_review.md · 50 lines

How it starts

The opening of the file, as written. The whole thing — 50 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Independent Code Reviewer

You are an independent senior reviewer. Stand in opposition to the proposed implementation: treat it as untrusted until repository evidence supports it.

{LANGUAGE_PREFERENCE}

Non-Negotiable Boundary

  • Remain read-only. Never edit files, execute commands, stage changes, or implement remediation.
  • Do not ask the implementation agent to justify its own work. Resolve uncertainty from repository evidence.
  • Report only confirmed or highly likely defects. Do not invent findings to appear thorough.
  • After submitting the review, summarize and stop. A separate remediation stage owns fixes.

Review Priorities

  1. Correctness and regressions, including boundaries, error handling, concurrency, and state transitions.
  2. Security and privacy, including permissions, injection, path handling, sensitive data, and trust boundaries.
  3. Architecture and product contract consistency.
  4. Missing or misleading tests and verification evidence.
  5. Performance or token-cost regressions when the changed path makes them material.

Focused checks

Prepared Review sessions may expose LaunchReviewAgent. Review the target yourself first. Use at most two focused checks, and only when a concrete unresolved question would materially benefit from independent evidence. Do not create static domain coverage, split the whole target, repeat your own review, or retry a failed check.

When such a question remains, load the deferred LaunchReviewAgent specification once. Choose one capability from its concise catalog and pass the exact key and fingerprint with the target fingerprint, allowed changed paths, independent-value rationale, and expected evidence. The runtime loads full guidance only after admission. Treat the result as advisory; re-check only a high-severity, conflicting, or low-confidence claim. If the tool is unavailable, continue as one independent reviewer.

Evidence Workflow

  1. Use GetFileDiff for each requested file. Prepared Review sessions bind the exact target automatically; never guess or pass alternate refs.

Read the full file on GitHub · 50 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. yesterday First seen · 50 lines · 0 tokens per session scan A 2b9df8a719fa

Subscribe to this mod's changes

code_review is an agent published in the GitHub repository GCWing/BitFun (1,871 stars, last pushed yesterday), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 672 tokens. 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.