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 Olorinm/visual-review-workbench --skill review-frontend-variantsgit clone --depth 1 https://github.com/Olorinm/visual-review-workbenchWrote 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/olorinm/visual-review-workbench/review-frontend-variants)<a href="https://agentmods.dev/skills/olorinm/visual-review-workbench/review-frontend-variants"><img src="https://agentmods.dev/badge/skills/olorinm/visual-review-workbench/review-frontend-variants/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/olorinm/visual-review-workbench/review-frontend-variants"><img src="https://agentmods.dev/badge/skills/olorinm/visual-review-workbench/review-frontend-variants.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 | $0.00690 |
| Opus 5.5 | $0.00034 | $0.00276 |
| Sonnet 5.5 | $0.00017 | $0.00138 |
| Haiku 4.5 | $0.00008 | $0.00069 |
Grade A, and why
review-frontend-variants 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 — 44 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Frontend Review Workbench
Give the user one compact review surface containing trustworthy UI sources. The workbench arranges evidence; it does not recreate product UI.
Allowed sources
Use the most faithful available source and label it clearly:
- Live — the ordinary route of a real running product. For Before/Candidate comparisons, run the real app from separate temporary worktrees and ports.
- Story — a Storybook, React Cosmos, or equivalent scenario already maintained by that product.
- Capture — a Playwright element/page screenshot or recording from the real product when live embedding is unavailable.
For external references, run each project in its own native environment. Never import several products into one bundle or imitate their surrounding UI. If a product cannot run or be embedded, show a clearly marked capture instead.
Workflow
- Confirm the exact screen, state, data, theme, locale, and viewport to compare.
- Start the real product instances. Use separate temporary worktrees for incompatible candidates; do not add query-driven variants to product code.
- Prepare the state through real product UI/API or product-maintained seeds and test factories. Keep the recipe task-local, deterministic, and idempotent; use it unchanged for Before and Candidate.
- Create the review config and generated output outside product repositories. Point panels at real URLs, native stories, or real captures.
- Add action assertions for the expected visible state, then probe and inspect the completed workbench.
- Run
ui-review audit --repo <product> --base <target-branch>against every product repository before delivery; never use the feature branch'sHEADas the baseline. After selection, keep only the chosen production implementation and behavior tests in the product PR.
Hard boundaries
- Never add or commit review-only routes, fixtures, fake bootstrap/state, no-op services, copied product shells, variant readers, or review CSS to a product repository.
- Never invent state through request interception, frontend-store mutation, or hard-coded review data. If a real workflow or native seed is impractical, use an existing Story or a real Capture and label it accordingly.
- Never put personal absolute paths in tracked files. Runtime paths belong only in task-local, untracked config.
- Never call an isolated atom a complete product screen. Preserve its real parent context or label it as a component Story.
- Do not modify reference repositories. Run them read-only or capture them.
- Do not add external references unless the user requested them.
- Do not leave rejected candidates or review infrastructure in the final product diff.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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 Changed · -196 lines · -15 tokens per session 24712ce22570
- 1mo ago First seen · 240 lines · 99 tokens per session scan A 5f3feb3f8417
review-frontend-variants is a skill published in the GitHub repository Olorinm/visual-review-workbench (0 stars, last pushed 1mo ago), licensed MIT. It adds 84 tokens to every session and 690 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 2026-08-31.
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.