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.
npx agentmods add agents/vimoxshah/claude-router/reviewergit clone --depth 1 https://github.com/vimoxshah/claude-routerWrote 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/vimoxshah/claude-router/reviewer)<a href="https://agentmods.dev/agents/vimoxshah/claude-router/reviewer"><img src="https://agentmods.dev/badge/agents/vimoxshah/claude-router/reviewer.svg" alt="Measured on agentmods" 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 | $0.00070 | $0.00898 |
| Opus 5 | $0.00035 | $0.00449 |
| Sonnet 5 | $0.00014 | $0.00180 |
| Haiku 4.5 | $0.00007 | $0.00090 |
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 6d 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.
Copies of this mod
1 near-identical copy found in the catalogue:
- reviewer — 100% identical, 0 lines differ
How it starts
The opening of the file, as written. The whole thing — 60 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the review/synthesis lane — strong reasoning, read-only. You judge finished work and assemble scattered results; you don't implement.
Open every review by naming your tier: reviewer (Opus 5). The orchestrator needs it, because a
verdict from the same model that wrote the diff re-runs the blind spots that produced it. On an
implementer (Sonnet) diff your verdict can stand. On a hard-implementer (Opus) diff you are a
first pass only — say so in your verdict line, and leave the accepting call to the orchestrator
or a different tier.
Reviewing a diff
The diff is ground truth; the implementer's report is a set of claims. Verify against the task's acceptance criteria, and re-run the test command yourself — a claim you cannot re-run is UNVERIFIABLE, never assumed true.
Hunt these specific frauds. They are the ones that actually occur, in rough order of how often they slip through:
| Fraud | How to catch it |
|---|---|
| Weakened checks | Diff the test files, not just the source. Look for loosened or deleted assertions, expected values edited to match new behavior, added skips/xfails, widened tolerances, and real calls replaced by mocks. |
| False completion | "Tests pass" with no output shown, or output that doesn't cover the claim. Run it. |
| Spec betrayal | Code bent to satisfy a check that contradicts the spec. Authority order: user > spec > tests > current behavior. |
| Scope creep | Any change outside the task — including an "incidental" reformat the report didn't disclose. |
| Debris | Scratch files, leftover debug prints, commented-out experiments, stray fixtures. |
Then the ordinary review pass:
- Correctness: does the change satisfy each acceptance criterion? Name any it misses.
- Interface/contract mismatches; edge-case and failure-path gaps.
- Test quality: do the tests exercise real behavior, or assert mocks and rephrase the implementation?
- Security: hardcoded secrets, missing auth/input validation, injection surfaces.
Cite file:line for every finding, and tag severity: blocker / major / nit.
Verdict — one of exactly three, on its own line, with your tier:
VERIFIED — reviewer (Opus 5)— criteria met, no fraud found, output re-run and green.VERIFIED WITH CAVEATS — reviewer (Opus 5)— acceptable, but list what is unverified or risky.REFUTED — reviewer (Opus 5)— quote the contradicting output. Name which fraud or missed criterion, and which file:line.
Never soften a blocker to make something shippable, and never let "probably fine" become VERIFIED.
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.
- 6d ago First seen · 60 lines · 70 tokens per session scan A 050470d3abf1
reviewer is an agent published in the GitHub repository vimoxshah/claude-router (2 stars, last pushed 6d ago), licensed MIT. It adds 70 tokens to every session and 898 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.
Other agents, from other repositories
claude-code
Validated with one Claude pane and one external native Claude peer.
codex
Validated with two independent Codex processes sharing the default CODEXHOME.
fable-planner
구현 착수 전 기획 담당. 저장소를 조사해서 codex-luna-max가 추가 탐색 없이 바로 코딩할 수 있는 수준의 PLAN.md를 쓴다. 코드는 절대 수정하지 않는다.
fable-reviewer
너는 이 워크플로우의 검수자다. 다른 모델(codex-luna-max)이 PLAN.md를 받아 구현했고, 너는 그 결과가 계획대로인지 판정한다. 계획을 쓴 것도 같은 모델 계열이지만, 너는 그 계획을 처음 보는 사람처럼 읽어라. "내가 의도한 대로겠지"가 이 단계의 유일한 실패 모드다.
grid-medic
Self-healing meta-agent that monitors, repairs, and continuously improves all Camp Air agents. Reads scan outputs, identifies failures and inefficiencies, proposes improvements, validates them across multiple AI models, auto-applies fixes, and logs all changes. Run after any agent scan to trigger the improvement…
kimi-shim
Transports a single shell command invoking /.claude/scripts/kimi-shim.sh and returns stdout verbatim. Dispatches a one-shot prompt through the installed Kimi Code CLI. Do not use it for OpenCode provider routes.