SkillsBench is a benchmark for measuring how effectively AI agents use modular skills—folders containing instructions, scripts, and resources—to complete specialized tasks. It helps researchers and developers evaluate both skill quality and agent behavior, including tasks that require combining multiple skills. The catalogue’s skills and instructions are evaluated as part of this workflow.
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 benchflow-ai/skillsbench --skill ada-plan-view-accessibilitygit clone --depth 1 https://github.com/benchflow-ai/skillsbenchWrote 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/benchflow-ai/skillsbench/ada-plan-view-accessibility)<a href="https://agentmods.dev/skills/benchflow-ai/skillsbench/ada-plan-view-accessibility"><img src="https://agentmods.dev/badge/skills/benchflow-ai/skillsbench/ada-plan-view-accessibility/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/benchflow-ai/skillsbench/ada-plan-view-accessibility"><img src="https://agentmods.dev/badge/skills/benchflow-ai/skillsbench/ada-plan-view-accessibility.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00041 | $0.00805 |
| Opus 5.5 | $0.00016 | $0.00322 |
| Sonnet 5.5 | $0.00008 | $0.00161 |
| Haiku 4.5 | $0.00004 | $0.00081 |
Grade A, and why
ada-plan-view-accessibility 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 — 49 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Use this skill for plan-view accessibility checks. It is not a full legal ADA compliance review; apply only the geometric rules provided by the task.
Plan-View Checks
- Door clear opening: compare the submitted clear width against the minimum in the rules file, commonly 32 inches.
- Door swing: if the rule disallows swing into required fixture clearance, prefer an outward or sliding swing in the repaired layout. Do not leave an inward repaired swing unless you have explicitly modeled and cleared the swing path against fixtures and required clearances.
- Turning space: verify that a 60 inch diameter circle fits inside usable floor area. Diameter alone is not enough. Apply the wall/boundary offset from the rules, subtract protected or blocked fixtures such as tubs, and allow overlap with toilet clear floor space only when the rules say so.
- Lavatory overlap: allow the turning circle to overlap a lavatory only when plan-view knee/toe clearance is explicitly indicated and meets the required width and depth.
- Toilet centerline: measure from the adjacent side wall and keep it inside the rule range, commonly 16 to 18 inches.
- Grab bars: classify rear-wall and side-wall bars by orientation/location and compare their plan-view lengths against the minimums. The side-wall grab bar should align with the same adjacent side wall used for the toilet centerline and overlap the toilet use zone. The rear-wall grab bar should align with the rear wall behind the toilet and cross or closely span the toilet centerline.
Violation Reporting
Report one violation per failed rule/element pair. Use stable machine-readable rule names from the rules file, for example:
door_clear_width_minturning_circle_diameter_minturning_circle_fit_usable_floortoilet_centerline_from_side_wall_rangeside_grab_bar_length_minrear_grab_bar_length_minside_grab_bar_toilet_side_wall_alignmentrear_grab_bar_toilet_centerline_spanlavatory_knee_clearance_min
Do not report vertical requirements such as seat height, mounting height, mirror height, signage height, or pipe protection if the task scope excludes them.
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 · 49 lines · 41 tokens per session scan A d6177e0345a6
ada-plan-view-accessibility is a skill published in the GitHub repository benchflow-ai/skillsbench (1,871 stars, last pushed 2d ago), licensed Apache-2.0. It adds 41 tokens to every session and 805 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 2026-08-30.
Other skills, from other repositories
accessibility-compliance-accessibility-audit
Audit pages and user journeys against a declared WCAG scope. Use when identifying keyboard, screen-reader, focus or contrast barriers and planning verified remediation.
counsel
Classic, trustworthy design for law firms, attorneys and notaries with navy, ivory and brass colors, serif headings, sharp corners and fine gold rules.
mqc-litigation-visual-redraw
Redraw a litigation diagram into a restrained, court-ready presentation graphic (SVG + PNG) WITHOUT changing any text or legal meaning. Use this whenever the user supplies a case timeline, a legal process flowchart, OR a party/relationship diagram and wants it cleaned up, beautified, redrawn, made professional…
mqc-legal-relation-master
A legal case-preparation workflow that turns case materials into a diagram of people, contracts, projects, money, and the relationships between them. It produces editable diagram files as well as image versions, using the selected visual style.
arckit-at-barrierefreiheit-tractorjuice
You are helping an enterprise architect generate an Austrian Accessibility Assessment. Austria transposed the EU accessibility regime along two separate tracks, and which one applies is a question about the entity, not about the technology.
auditing-wcag
Perform formal WCAG 2.2 A/AA conformance audits and assign Pass/Fail/NT/NA to every success criterion with evidence. Use reviewing-a11y instead for issue discovery and improvement feedback.