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 marcoguillermaz/Tierward --skill visual-auditgit clone --depth 1 https://github.com/marcoguillermaz/TierwardWrote 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/marcoguillermaz/tierward/visual-audit)<a href="https://agentmods.dev/skills/marcoguillermaz/tierward/visual-audit"><img src="https://agentmods.dev/badge/skills/marcoguillermaz/tierward/visual-audit/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/marcoguillermaz/tierward/visual-audit"><img src="https://agentmods.dev/badge/skills/marcoguillermaz/tierward/visual-audit.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 | $0.00000 | $0.05167 |
| Opus 5 | $0.00000 | $0.02583 |
| Sonnet 5 | $0.00000 | $0.01033 |
| Haiku 4.5 | $0.00000 | $0.00517 |
Grade B, and why
visual-audit scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Recursive force deletemediumDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
rm -rf "${SCREENSHOT_TEMP_DIR:-/tmp/visual-audit-screenshots}" Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
How it starts
The opening of the file, as written. The whole thing — 406 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Configuration (fill in before first run)
Replace these placeholders:
[SITEMAP_OR_ROUTE_LIST]- e.g.docs/sitemap.md,docs/routes.md,src/router/index.ts[DEV_URL]- e.g.http://localhost:3000,http://localhost:5173[DEV_COMMAND]- e.g.npm run dev,pnpm dev[TEST_ACCOUNTS]- reference to test credentials (e.g. "see CLAUDE.md Environment section")[LOGIN_ROUTE]- e.g./login,/sign-in,/authIf
[SITEMAP_OR_ROUTE_LIST]or[DEV_URL]is not filled, the skill reports an error and exits.
Scope boundary: this skill covers aesthetic quality only. Accessibility contrast (APCA measurement, WCAG contrast thresholds, axe-core scan) lives in /accessibility-audit - run it for any contrast or accessibility concern. Design-token compliance → /ui-audit. Viewport fit → /responsive-audit.
Step 0 - Mode + target detection
Parse $ARGUMENTS:
Mode (controls page roster breadth):
quick→ 5 key pages onlyfull→ all routes from sitemap, both themes- No mode keyword → standard - all routes from
[SITEMAP_OR_ROUTE_LIST](light + dark) critique→ deep-dive on a singletarget:page:(required). Skips scoring table; produces Gestalt + hierarchy diagnosis + concrete redesign proposals per Critical/Major finding. Must be paired withtarget:page:<route>.
Target (filters the page roster):
target:role:<role_name>→ routes accessible to that role onlytarget:section:<name>→ routes whose path contains<name>- No target → NO filter - use ALL pages from the mode roster per
[SITEMAP_OR_ROUTE_LIST]
Target scoping - comparative vs independent checks (mandatory): some audit skills contain comparative checks that derive a project-wide convention by counting usage across the whole inventory - those must ignore the target: filter. This skill has none: every dimension is scored per page against the absolute anchors in DIMENSIONS.md, never relative to other pages.
- Comparative checks - full-project by design; the
target:filter does NOT apply to them: none. No check derives its pass/fail baseline by counting across the page inventory. - Independent checks - target-safe: all checks (Step 5a code inspection, Step 5c computed checks, dimensions V1-V7 and V9-V11) evaluate one page in isolation and safely honor the
target:filter. - When a
target:is given: apply the filter to the working roster for all steps. Step 7 (cross-page pattern analysis) then aggregates only the audited subset - its "systemic issues", "best-in-class", and "worst offenders" labels are scope-relative under a target, and the report MUST qualify them as such rather than present them as project-wide claims.
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.
- 9d ago First seen · 406 lines · 0 tokens per session scan B d2b88dfa0855
visual-audit is a skill published in the GitHub repository marcoguillermaz/Tierward (4 stars, last pushed 2d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 5,167 tokens. A static security scan graded it B with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
frontend-taste
Premium frontend design skills that override LLM defaults. Prevents generic 'AI-looking' UI. Enforces deterministic typography, calibrated color, asymmetric layouts, spring physics, hardware-accelerated motion. Triggers: 'UI tasarla', 'design a UI', 'premium landing page', 'frontend taste', 'anti-slop', 'dashboard…
design-tokens
Project-level DESIGN.md tokens reference. When the project has a DESIGN.md file, agents producing UI / components / visuals must consult this skill to use canonical color/typography/spacing tokens rather than inventing new ones. Triggers on: UI generation, component creation, design brief, visual asset, tailwind…
design-taste-frontend
Anti-slop frontend skill for landing pages, portfolios, and redesigns. The agent reads the brief, infers the right design direction, and ships interfaces that do not look templated. Real design systems when applicable, audit-first on redesigns, strict pre-flight check.
design-taste-frontend-v1
The original v1 taste-skill, preserved for projects depending on its exact behavior. The current default is design-taste-frontend (v2 experimental), which is a substantial rewrite. Use this v1 install name only if you need exact backward compatibility.
high-end-visual-design
Teaches the AI to design like a high-end agency. Defines the exact fonts, spacing, shadows, card structures, and animations that make a website feel expensive. Blocks all the common defaults that make AI designs look cheap or generic.
gpt-taste
Elite UX/UI & Advanced GSAP Motion Engineer. Enforces Python-driven true randomization for layout variance, strict AIDA page structure, wide editorial typography (bans 6-line wraps), gapless bento grids, strict GSAP ScrollTriggers (pinning, stacking, scrubbing), inline micro-images, and massive section spacing.