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 skills add MadAppGang/magus --skill review-servicesgit clone --depth 1 https://github.com/MadAppGang/magusWrote 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/skills/madappgang/magus/review-services)<a href="https://agentmods.dev/skills/madappgang/magus/review-services"><img src="https://agentmods.dev/badge/skills/madappgang/magus/review-services/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/skills/madappgang/magus/review-services"><img src="https://agentmods.dev/badge/skills/madappgang/magus/review-services.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 | $0.00048 | $0.02248 |
| Opus 5.5 | $0.00019 | $0.00899 |
| Sonnet 5.5 | $0.00010 | $0.00450 |
| Haiku 4.5 | $0.00005 | $0.00225 |
Grade A, and why
review-services 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 1mo 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 — 193 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Review services
Two procedures the designer agents share. Procedure A is how a screen gets judged by a model other than the one running the agent. Procedure B is how the agent finds out which review services this project already pays for, and uses them when they are there.
Both procedures write what they did into the report header, so a reader always knows which eyes looked at the screen and which scanners ran.
Procedure A — external vision judge
Why an external model
The agent running this skill can Read a PNG and see it. That is a second opinion from
the same vendor. The owner's rule for design judging is a different vendor's eyes: the
newest Gemini Pro the catalog lists, and the top GPT tier when no Gemini Pro is live.
Model IDs are never written down here; they are resolved at run time.
Step 1 — resolve the judge
search_models(query="gemini")
Pick the entry whose id names a Gemini Pro model with the highest version. Ignore
Flash, Lite, image-generation and preview-only variants unless no Pro entry exists.
If no Gemini Pro entry is live:
list_models()
Pick the highest-version gpt-* entry in the top tier. A user-named version is a hard
constraint: use it if listed, otherwise stop and name the live alternatives. Never
downgrade to a lower version because its name is closer as a string.
Send the catalog's id, never an openrouterId or a provider@model address.
Record JUDGE_MODEL = <id> for the report header.
Step 2 — write the brief
The child session reads the images itself; claudish carries the image blocks to the provider. So the brief names paths, and the paths must be readable from the working directory the child runs in (the current one).
Write ${JUDGE_DIR}/input.md, where JUDGE_DIR is a directory inside the current
working directory (${OUTPUT_DIR}/judge when OUTPUT_DIR is inside it; otherwise
.ui-validation/judge-<RUN_ID>). The team tool refuses a path outside the working
directory.
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.
- 1mo ago First seen · 193 lines · 48 tokens per session scan A 543c61d3cef9
review-services is a skill published in the GitHub repository MadAppGang/magus (13 stars, last pushed today), licensed MIT. It adds 48 tokens to every session and 2,248 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-09-15.
Other skills, from other repositories
webgl-holographic-foil
A self-contained WebGL2 hero: thin-film interference over a crushed-foil surface whose palette shifts with the viewing angle; move the cursor to tilt the film.
html-ppt-hermes-cyber-terminal
OpenDesign + BYOK: choosing and wiring your own model, hands-on — cost, quality, and the routing decision. Built as a decision-grade AI literacy deck for engineers, IT, applied-AI teams.
html-ppt-taste-brutalist
16:9 HTML deck in tactical-telemetry / CRT-terminal taste. Deactivated-CRT charcoal slides, white-phosphor monospace, hazard-red accent, scanline overlay, ASCII syntax, density over decoration. Distilled from Leonxlnx/taste-skill brutalist-skill (Tactical Telemetry mode).
visual-ralph
Visual Ralph orchestration for frontend UI from generated references, static references, or live URL targets, using $ultragoal with built-in visual verdict and pixel-diff evidence until the implementation matches and leaves a reproducible design system.
glance-canonical-layouts
Use this skill to create home screen widgets using Jetpack Glance Canonical Layouts from the Android Platform Samples project, or when asked to implement standard widget layout patterns (including Check List, Image Grid, Snap Scrolling, and Standard Toolbar).
accessibility
Consolidated accessibility skill entrypoint for WCAG 2.2, ARIA Authoring Practices, cognitive accessibility, Section 508, EN 301 549, design intent verification, and the Accessibility Planner workflow.