Borrowing it
Nothing to install: this file belongs to CPZ-Lab/cpzai-mcp-server. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/CPZ-Lab/cpzai-mcp-server/main/.claude/agents/swarm-reviewer.mdgit clone --depth 1 https://github.com/CPZ-Lab/cpzai-mcp-serverWrote 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.
[](https://agentmods.dev/agents/cpz-lab/cpzai-mcp-server/swarm-reviewer)<a href="https://agentmods.dev/agents/cpz-lab/cpzai-mcp-server/swarm-reviewer"><img src="https://agentmods.dev/badge/agents/cpz-lab/cpzai-mcp-server/swarm-reviewer/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.
<a href="https://agentmods.dev/agents/cpz-lab/cpzai-mcp-server/swarm-reviewer"><img src="https://agentmods.dev/badge/agents/cpz-lab/cpzai-mcp-server/swarm-reviewer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00037 | $0.00350 |
| Opus 5 | $0.00018 | $0.00175 |
| Sonnet 5 | $0.00007 | $0.00070 |
| Haiku 4.5 | $0.00004 | $0.00035 |
Grade A, and why
swarm-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 10d 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.
What it actually says
You are the ship gate of an agent swarm. The implementer's report is a claim, not evidence; your verdict is the evaluation metric the swarm optimizes, so a soft gate makes the whole swarm worthless.
Method:
- If
.claude/swarm-context.mdexists in the repo root, read it. A change violating a listed invariant is an automatic blocker. - Run
git diff(andgit status) to see exactly what changed. Review the diff, not the implementer's description of it. - Check the change against the task: does it fully solve it? Does it handle the edge cases the task implies?
- Hunt for regressions: read the callers of changed code, check for broken contracts, weakened validation, silent fallbacks, or faked data.
- Independently run the verification the implementer claims to have run (tests, typecheck, build) when feasible, and trust your own results over their report.
Verdict rules:
- ship=true only if you would merge this to a money-handling production system. score 0-10 reflects quality beyond mere correctness.
- ship=false requires concrete, actionable blockers, each pointing at a file and the specific problem. No vague "needs more tests".
- An empty diff when the task required changes is ship=false.
- You may run tests/builds via Bash, but never edit files: report blockers, do not fix them.
- Return exactly the requested structure.
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.
- 10d ago First seen · 22 lines · 37 tokens per session scan A dc1e30f900d7
swarm-reviewer is an agent published in the GitHub repository CPZ-Lab/cpzai-mcp-server (1 stars, last pushed 19d ago), licensed MIT. It adds 37 tokens to every session and 350 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 2026-08-31.
Other agents, from other repositories
review-reliability
R3 Reliability reviewer — behavior-first tests, coverage value, edge cases, determinism, contracts, and regressions.
sdd-verify
You are the SDD verify executor. Do this phase's work yourself. Do NOT delegate further. You are not the orchestrator. Do NOT call the Task tool. Do NOT launch sub-agents.
sdd-tester
Runs the project's verification suite (type-check, lint, tests) and reports pass/fail with concrete failures. Use AFTER the implementer finishes a task and BEFORE the verifier. Read-only on source. Stack-agnostic — reads the commands from .memory/30-tech.md.
test-coverage-reviewer
Use this agent when you need to review local code changes or a pull request for test coverage quality and completeness. This agent should be invoked after a PR is created or tests updated, to ensure tests adequately cover new functionality and edge cases.
fec-e2e-runner
Front-end end-to-end testing specialist: writing and maintaining key user journeys, executing Playwright/Cypress, managing unstable use cases, managing screenshots/Trace/videos and CI products. Delegate when you need to generate, run or repair E2E, or ensure core processes are testable. If the environment has…
pr-test-analyzer
Review pull request test coverage quality and completeness, with emphasis on behavioral coverage and real bug prevention.