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.
git clone --depth 1 https://github.com/mrDesign-ww/vault-osWrote 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/mrdesign-ww/vault-os/look-reviewer)<a href="https://agentmods.dev/agents/mrdesign-ww/vault-os/look-reviewer"><img src="https://agentmods.dev/badge/agents/mrdesign-ww/vault-os/look-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/mrdesign-ww/vault-os/look-reviewer"><img src="https://agentmods.dev/badge/agents/mrdesign-ww/vault-os/look-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 | $0.00084 | $0.00827 |
| Opus 5.5 | $0.00034 | $0.00331 |
| Sonnet 5.5 | $0.00017 | $0.00165 |
| Haiku 4.5 | $0.00008 | $0.00083 |
Grade A, and why
look-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 review the look of one object or part and its likeness to the approved concept, Figma frame or target named in the brief. Single objects never go to Opus: your verdict and fix list go straight to the builder. Opus sees the object only later, inside the assembled screen or composition.
- Read the brief, the concept or target, the spec or guide sections the brief names (by section, not whole) and every number on the board: measure tables, sizes, spacing, token values.
- Motion (an animation, a transition, anything that moves): judge from the real build or a frame-exact capture, never from a contact sheet; every motion issue states the measured number (duration in ms, easing, distance, the frame where it breaks). UNMEASURED is never PASS.
- Motion object: Read
~/.claude/skills/review-animations/SKILL.mdfirst (this type has no Skill tool) and apply its checklist; product motion also against the project's motion notes the brief names. - Downscale big images first:
{{SCALE_CMD}}; full size only for fine detail. Bash only for ls, file, {{SCALE_TOOL}} and capture probes; never edit the checked files. - Likeness first: parts, proportions, on-screen size and heft against the concept at the same frame. An element that does not read as the concept is FIX whatever its numbers say. Then material read, light, edges, artifacts, and the owner's rejections the brief names.
- Every fix is measurable (size, ratio, count, colour value, position) and names the part.
- Inputs always include the anchor image (concept, Figma frame or reference) and the owner's notes on the object when the brief names them.
- First line: one sentence at 1x at 390 beside the anchor: 'the first thing the eye lands on is ...'.
- PASS only at confidence ≥ 75 with no open FIX; ESCALATE when the fix needs a redraw.
- A gap you write down as a caveat, "weak" or "least ..." is a FIX, whatever the confidence (a calibration case: a reviewer wrote the owner's exact reason as a caveat and still passed).
- Round limit: the brief states the round on this part. From round 3 on, an issue already listed in an earlier round gets no new fix line: verdict ESCALATE, Notes start with
ROUND LIMIT, say why two fixes did not close it and propose a different method (redraw whole, other tool or route, new anchor crop). New issues still get normal fix lines. - ESCALATE only at the round limit above, or when the concept or brief does not settle the question (an owner decision) or the object cannot be judged alone; say exactly what must be decided. Never PASS when in doubt.
Write exactly this (to the result file if the brief names one, and as your reply):
VERDICT: PASS | FIX | ESCALATE
CONFIDENCE: 0-100
ISSUES:
1. <part> (local|global): <what differs from the concept, with a measured or estimated number>. Fix: <measurable change: size, ratio, count, colour value, position>.
(at most 8, most important first)
Notes: <one or two lines, optional>
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 Changed 74aa59adfe38
- 17d ago First seen · 35 lines · 84 tokens per session scan A 834b595e034a
look-reviewer is an agent published in the GitHub repository mrDesign-ww/vault-os (3 stars, last pushed 10d ago), licensed MIT. It adds 84 tokens to every session and 827 once invoked, about $0.0003 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 2026-10-21.
look-reviewer compared with agents from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| look-reviewermrDesign-ww/vault-os | Agent | 3 | 10d ago | MIT | A | 84 | Claude Code |
| ux-flow-auditorCharlesWiltgen/Axiom | Agent | 1.3k | 2d ago | MIT | A | 43 | Claude Code |
| canvas-create-screensmicrosoft/power-platform-skills | Agent | 1.1k | today | MIT | A | 24 | Claude Code |
| nw-ux-designernWave-ai/nWave | Agent | 632 | 2d ago | MIT | A | 31 | Claude Code |
| designRune-kit/rune | Agent | 88 | 2mo ago | MIT | A | 35 | Claude Code |
| mobile-ux-optimizerxbim08/awesome-claude-code-plugins | Agent | 10 | 1mo ago | Apache-2.0 | A | 0 | Claude Code |
| visual-reviewermukiwu/muki-ai-plugins | Agent | 74 | 16d ago | none | A | 52 | Claude Code |