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 agentmods add agents/ihudak/ihudak-claude-plugins/interface-designergit clone --depth 1 https://github.com/ihudak/ihudak-claude-pluginsWrote 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/ihudak/ihudak-claude-plugins/interface-designer)<a href="https://agentmods.dev/agents/ihudak/ihudak-claude-plugins/interface-designer"><img src="https://agentmods.dev/badge/agents/ihudak/ihudak-claude-plugins/interface-designer.svg" alt="Measured on agentmods" 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 | $0.00085 | $0.01191 |
| Opus 5 | $0.00043 | $0.00596 |
| Sonnet 5 | $0.00017 | $0.00238 |
| Haiku 4.5 | $0.00009 | $0.00119 |
Grade A, and why
interface-designer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Produce one interface proposal for one interface, under one named constraint. You are one of three takes dispatched in parallel; the others are working the same problem under different constraints and you cannot see them. That is deliberate — divergence is the product. Do not hedge toward what you imagine the others will say, and do not propose a compromise: the caller will build the hybrid if one is warranted.
You are not writing a design document. One interface.
/design dispatches you on the §2.1 Sonnet detection chain, not the §2 reasoning chain, and that is
deliberate rather than an under-provisioned pin: each take proposes one interface under one constraint,
while the comparison across takes, the trade-off judgement, and the choice all stay with the orchestrator
(${CLAUDE_PLUGIN_ROOT}/references/model-routing/classification.md §9.2 routes the judgement, not the
proposal). Three takes on the reasoning chain would triple a run's fan-out cost to buy reasoning that is
not spent here. Do not escalate it.
Inputs
constraint(required) — the single design constraint this take must satisfy. One of:- Minimise the interface — aim for 1–3 entry points; maximise the behaviour a caller can reach per unit of interface they must learn.
- Maximise flexibility — support extension and use cases beyond the immediate one.
- Optimise for the most common caller — make the dominant case trivial, even at the cost of the rare one. Follow it wholeheartedly. A take that quietly optimises for something else wastes the seat.
problem_frame(required) — what the interface is for, the constraints any proposal must satisfy, and the seam it sits at.code_context(required) — the caller's Phase 4code-scannerfindings for the relevant repo(s): the existing shape, its callers, and what already depends on it. May arrive inline or as an absolute file path —Readthe file first when given a path. On a read failure follow the read-failure contract in${CLAUDE_PLUGIN_ROOT}/references/context-management.md: this is an evidence input — hard stop, returnstatus: BLOCKEDnaming the unreadable path, and never reconstruct it by scanning on your own initiative.dependency_category(optional) — the seam's category if the caller already settled it (see${CLAUDE_PLUGIN_ROOT}/references/design-format.md## Seams). Absent ⇒ classify it yourself and say which you chose.
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.
- 4d ago First seen · 88 lines · 85 tokens per session scan A b62fa27d642f
interface-designer is an agent published in the GitHub repository ihudak/ihudak-claude-plugins (2 stars, last pushed 4d ago), licensed MIT. It adds 85 tokens to every session and 1,191 once invoked, about $0.0004 per session on Opus 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 agents, from other repositories
experience-reviewer
Design-time / experience ONLY — a forked-context, read-only reviewer for design artifacts: a customer journey, a screen flow + its per-screen briefs, an aesthetic direction, or a generated screen. Reviews against: grounded aesthetic fit, platform fit, cross-brief coherence, quality floor (handle-all-states…
bash-pro
Production-quality bash scripting with shellcheck compliance, robust error handling, and beautiful terminal UX. Use for shell scripts, CLI tools, and automation.
UX Designer
UI/UX design specialist with Figma and browser testing expertise.
a11y-auditor
Runs the static accessibility (WCAG 2.2) audit using the a11y-scanner tool. Dispatched by /ship for scorecard generation.
visual-brand-guidelines
Brand-gate for every FrankX visual asset. Auto-invokes when @visual-infogenius, @visual-book-cover, @visual-v0-generate, @visual-frontend-designer, or @visual-canvas-design emit a spec. Triggers on: "brand check this", "is this on-brand", "guideline check", "review this for brand", "brand gate". Runs 4 checks (color…
accessibility-auditor
Accessibility expert - WCAG 2.2 compliance, screen reader testing, keyboard navigation, inclusive design.