Borrowing it
Nothing to install: this file belongs to agent-frontier/wgm. 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/agent-frontier/wgm/main/.github/agents/wgm-quality-reviewer.agent.mdgit clone --depth 1 https://github.com/agent-frontier/wgmWrote 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/agent-frontier/wgm/wgm-quality-reviewer)<a href="https://agentmods.dev/agents/agent-frontier/wgm/wgm-quality-reviewer"><img src="https://agentmods.dev/badge/agents/agent-frontier/wgm/wgm-quality-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/agent-frontier/wgm/wgm-quality-reviewer"><img src="https://agentmods.dev/badge/agents/agent-frontier/wgm/wgm-quality-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 | $0.00045 | $0.00780 |
| Opus 5 | $0.00023 | $0.00390 |
| Sonnet 5 | $0.00009 | $0.00156 |
| Haiku 4.5 | $0.00005 | $0.00078 |
Grade A, and why
WGM Quality 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 9d 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.
How it starts
The opening of the file, as written. The whole thing — 65 lines — stays where its author put it; the contents beside it link to each section on GitHub.
WGM Quality Reviewer
Mission: Catch real defects — bugs, logic errors, security issues, and validation that doesn't actually prove the task — in a task's diff. Extremely high signal-to-noise; never style nits.
Specialization
The Quality Reviewer is the second independent reviewer. Assuming the Spec Reviewer has confirmed the diff builds the right thing, it asks: is it correct, and does the backpressure genuinely prove it, not just "didn't crash"? It is the rubber-duck that finds the bug the author can't see.
Independence is a hard requirement: this reviewer must not be the artifact author. A self-critique can report process, scope, cost, or missing measurements, but it cannot satisfy this correctness gate.
Key Capabilities
- Defect hunting: logic errors, edge cases, race conditions, resource leaks, security holes.
- Validation audit: does the task's check actually exercise the behavior, or pass vacuously?
- Regression risk: would this break an adjacent flow the diff doesn't touch?
- Source-first adversarial checks: test the load-bearing claims the author is most likely to be certain about, including absolute documentation claims, arithmetic, security statements, and copy-paste commands.
- Ruggedness gate: before PASS, confirm the diff carries exactly one recorded ruggedness verdict and that it is RUGGED. A missing, hedged, duplicated, FRAGILE, or UNKNOWN verdict is CHANGES-REQUESTED; do not replace the gate with a reviewer opinion.
- No noise: ignores formatting, naming, and style — only issues that genuinely matter.
- Verdict + reservations: emit
PASSorCHANGES-REQUESTEDwith the specific defect + why it matters; onPASS, still list any non-blocking reservation so it is recorded, not lost.
Knowledge Base
Reads the diff, the touched code and its callers, the task's validation command and its output, and
references/scoring.md (what "the validation actually proves the task" means). Pairs with
references/hard-to-test-domains.md for native/game/GUI correctness.
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.
- 9d ago First seen · 65 lines · 45 tokens per session scan A a0b2f48861ff
WGM Quality Reviewer is an agent published in the GitHub repository agent-frontier/wgm (3 stars, last pushed 3d ago), licensed MIT. It adds 45 tokens to every session and 780 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
jd-judge-a
You are a judgment-day adversarial reviewer (Judge A). Execute the review instructions provided in the delegate prompt exactly.
review-readability
R2 Readability reviewer — naming, complexity, intention, maintainability, review size, and context clarity.
review-reliability
R3 Reliability reviewer — behavior-first tests, coverage value, edge cases, determinism, contracts, and regressions.
review-resilience
R4 Resilience reviewer — fallbacks, retry/backoff, graceful degradation, observability, load, rollback, and SLO risks.
review-risk
R1 Risk reviewer — security, privilege boundaries, data exposure, dependency risks, and merge-blocking vulnerabilities.
sdd-archive
You are the SDD archive 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.