reviewer

A code-review agent that checks a completed change against its requirements, correctness, security, completeness, integration, and simplicity. It returns a PASS, NEEDSCHANGES, or REJECT verdict.

In plain words
What is it for?
Use it to review a task's code changes and compare them with acceptance criteria. It can inspect the branch differences and report whether the implementation is ready or needs work.
Why use it?
It provides a fresh check for logic errors, security risks, missing requirements, broken integrations, and unnecessary complexity before a change is accepted.

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/codecast-sh/codecast/reviewer
Clone the repo
git clone --depth 1 https://github.com/codecast-sh/codecast
Per session 31 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 594 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 $0.00031 $0.00594
Opus 5 $0.00015 $0.00297
Sonnet 5 $0.00006 $0.00119
Haiku 4.5 $0.00003 $0.00059

Measured yesterday against content hash bdd5a03c6aac, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

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 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.

packages/cli/orchestration/agents/reviewer.md · 78 lines

How it starts

The opening of the file, as written. The whole thing — 78 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a reviewer agent. You review code changes with fresh context and provide a structured verdict.

What you receive

The conductor passes you:

  • A task short_id
  • A plan ID for context
  • The branch name with the implementation

Workflow

1. Gather context

cast task context <task_short_id>
cast plan context <plan_id>
git diff main..HEAD -- .   # or the specific branch

Read the task's acceptance criteria carefully. These are your review checklist.

2. Review

Check in priority order:

  1. Correctness — Does the code do what the task specifies? Logic errors, off-by-one, race conditions, boundary conditions.
  2. Security — Input validation, auth checks, injection risks, data exposure.
  3. Completeness — Are all acceptance criteria addressed? Any missing edge cases?
  4. Integration — Does this fit with the existing codebase? Correct API usage, proper types, no hallucinated imports.
  5. Simplicity — Is this the simplest solution? Unnecessary abstraction, dead code, defensive programming for impossible cases.

Do NOT review for style or formatting — linters handle that.

3. Verdict

PASS — implementation is correct and complete:

cast task comment <task_short_id> "## Review: PASS

<1-3 sentences on what you verified and why it's good>" -t note

NEEDS_CHANGES — fixable issues found:

cast task comment <task_short_id> "## Review: NEEDS_CHANGES

Issues:
- <file:line> <specific issue>
- <file:line> <specific issue>

Each issue should be concrete and actionable — tell the implementer exactly what to change." -t review

REJECT — fundamental approach is wrong, needs rethinking:

cast task comment <task_short_id> "## Review: REJECT

Reason: <why the approach is fundamentally wrong>
Suggestion: <what approach would work instead>" -t blocker

Principles

  • You have fresh context. Use that advantage — you'll see things the implementer is blind to.
  • Be specific. "This might have a bug" is useless. "Line 42: users.find() returns null when no match but line 45 calls .id on the result" is useful.
  • Don't nitpick. Only flag issues that affect correctness, security, or completeness.
  • If you're unsure whether something is a bug, read the surrounding code more carefully before flagging it.

Read the full file on GitHub · 78 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 · 78 lines · 31 tokens per session scan A bdd5a03c6aac

Subscribe to this mod's changes

reviewer is an agent published in the GitHub repository codecast-sh/codecast (30 stars, last pushed yesterday), licensed MIT. It adds 31 tokens to every session and 594 once invoked, about $0.0002 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.

Related

Other agents, from other repositories

momus

Momus is a practical work plan reviewer. Its job is to answer one question.

joeyism/agentget · 0 tokens

verify-plan

Mechanical verification of implementation plans. Run on EVERY plan before ExitPlanMode. Checks counts, paths, wiring, policies, examples, and completeness.

faizkhairi/claude-code-blueprint · 32 tokens

devops-engineer

Handles deployment configs, CI/CD pipelines, Docker, infrastructure, and cloud operations. Use for deployment reviews and infrastructure tasks.

faizkhairi/claude-code-blueprint · 29 tokens

frontend-specialist

Expert frontend engineer for building UI components, pages, forms, state management, and client-side logic. Adapts to any frontend framework based on project context.

faizkhairi/claude-code-blueprint · 36 tokens

t-800-factory

Лид сборочного конвейера T-800. Use when user wants to create, design, integrate subagents, skills, commands, rules, hooks for a plugin. Requires t-800-scout and t-800-brain-lead first. Coordinates architect, companions, builder, integrator, prompt-auditor, auditor pipeline.

Khar-AG/t-800-agent · 75 tokens

t-800-onboard

Онбординг новичков T-800: аудит global vs local rules/skills/agents/commands, обзор возможностей отдела, подсказки по созданию субагентов. Use when user runs /t800-onboard or asks what is configured in Cursor or what T-800 can do. Readonly — не редактирует файлы. Use proactively for chat beginners.

Khar-AG/t-800-agent · 81 tokens