evaluator

evaluator is an agent for Claude Code from anthropics/cwc-long-running-agents. It costs 63 tokens per session (391 once invoked), scanned A, original, Apache-2.0.

A second-opinion reviewer for software changes. It checks the requirements, code differences, screenshots, and logs, then returns either PASS or NEEDS_WORK.

In plain words
What is it for?
Use it to review completed features, inspect evidence, and report specific missing or broken acceptance criteria without editing files.
Why use it?
It reduces the risk of accepting a change based only on the builder’s claim that it works.

Agent for Claude Code ✓ vendor

Written for Claude Code: installed under .claude/.

Good fit Use it to review completed features, inspect evidence, and report specific missing or broken acceptance criteria without editing files.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/anthropics/cwc-long-running-agents/evaluator
About the project

anthropics/cwc-long-running-agents provides hooks, a subagent, and related files for building Claude Code workflows that keep working across long-running tasks. It is intended for developers who need agents to evaluate completion, preserve progress, and support operator intervention across sessions. The catalogue entries package these workflow primitives as hooks, an agent, and a setting.

anthropics/cwc-long-running-agents · 674 stars · on GitHub

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.

Clone the repo
git clone --depth 1 https://github.com/anthropics/cwc-long-running-agents

Made for: Claude Code.

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 evaluator

README.md
[![agentmods](https://agentmods.dev/badge/agents/anthropics/cwc-long-running-agents/evaluator.svg)](https://agentmods.dev/agents/anthropics/cwc-long-running-agents/evaluator)
Your own site
<a href="https://agentmods.dev/agents/anthropics/cwc-long-running-agents/evaluator"><img src="https://agentmods.dev/badge/agents/anthropics/cwc-long-running-agents/evaluator.svg" alt="Measured on agentmods" height="20"></a>
Per session 63 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 391 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 $0.00063 $0.00391
Opus 5 $0.00032 $0.00196
Sonnet 5 $0.00013 $0.00078
Haiku 4.5 $0.00006 $0.00039

Measured 8d ago against content hash 0ce0481cdf64, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

evaluator 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 8d 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.

claude-code-config/.claude/agents/evaluator.md · 26 lines

What it actually says

You are reviewing work that a separate builder agent just claimed is complete. You did not see how it was built and you should not trust the builder's own assessment.

Do the following every time:

  1. Read the spec or acceptance criteria for the feature under review.
  2. Run git diff against the baseline to see exactly what changed.
  3. Open every screenshot or console log under screenshots/ (or wherever the builder was told to put evidence) and look at what they actually show, not what the filenames imply. If a file fails to open or returns an error, treat it as missing evidence.
  4. Decide.

Plausibility is not correctness. A diff that looks reasonable paired with a screenshot that shows a broken layout is NEEDS_WORK. Missing evidence for any acceptance criterion is NEEDS_WORK. If you find yourself assuming something probably works, stop and look for proof.

Begin your reply with the bare word PASS or NEEDS_WORK on its own line, with nothing before it, so a wrapper script can read the verdict. Then:

  • PASS: one line stating what evidence convinced you.
  • NEEDS_WORK: a bullet list of specific, fixable findings the builder can act on next session.

Use Bash only for git diff, git log, and ls/cat. You cannot edit, write, or run the application. Do not offer to fix anything yourself.

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. 8d ago First seen · 26 lines · 63 tokens per session scan A 0ce0481cdf64

Subscribe to this mod's changes

evaluator is an agent published in the GitHub repository anthropics/cwc-long-running-agents (674 stars, last pushed 3mo ago), licensed Apache-2.0. It adds 63 tokens to every session and 391 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 2026-08-30.

Related

Other agents, from other repositories

reviewer

Read-only reviewer for an SDD implementation — checks that the change satisfies the acceptance criteria it claims (stage 1) and meets quality/convention/edge-case bars (stage 2). Use after a task (or the whole feature) reaches GREEN, before it's considered done. It reads the diff and the upstream artifacts and reports…

genkovich/sdd · 81 tokens

atomic-auditor

Final gate for a finished implementation. Dispatched exactly once after the implement-review loop goes green, never per iteration. Never touches the repo; its one write is the audit report into the task scratchpad. Audits the delivered work as a whole: cumulative spec compliance, cross-iteration coherence…

damusix/atomic-claude · 169 tokens

bt6-pr-auditor

Reviews one pull request in a BT6 codebase for correctness, research integrity, security, verification quality, and merge readiness.

elder-plinius/T3MP3ST · 32 tokens

Reviewer

Mandatory fast reviewer: validates every agent delegation output before acceptance. Checks acceptance criteria, file partitions, regressions, type safety, security basics.

monkilabs/opencastle · 30 tokens

security-auditor

Use this agent when reviewing local code changes or pull requests to identify security vulnerabilities and risks. This agent should be invoked proactively after completing security-sensitive changes or before merging any PR.

NeoLabHQ/context-engineering-kit · 40 tokens

dotnet-architecture-reviewer

Reviews a .NET codebase or repository and produces a structured architecture report — layering and dependency-rule violations, coupling, CQRS/handler hygiene, EF Core boundary leaks, testability, and concrete prioritized fixes. Use when the user wants an architecture review, a "second opinion" on structure, a PR-level…

StefanTheCode/dotnet-ai-toolkit · 102 tokens