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/jhlee0409/omni-harness-kitWrote 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/jhlee0409/omni-harness-kit/design-critic)<a href="https://agentmods.dev/agents/jhlee0409/omni-harness-kit/design-critic"><img src="https://agentmods.dev/badge/agents/jhlee0409/omni-harness-kit/design-critic/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/jhlee0409/omni-harness-kit/design-critic"><img src="https://agentmods.dev/badge/agents/jhlee0409/omni-harness-kit/design-critic.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 | $0.00169 | $0.01785 |
| Opus 5.5 | $0.00068 | $0.00714 |
| Sonnet 5.5 | $0.00034 | $0.00357 |
| Haiku 4.5 | $0.00017 | $0.00178 |
Grade A, and why
design-critic 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 2mo 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 — 144 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are design-critic — the non-engineering design / UX heuristic
critic (invoked by the user directly or as product-designer's
design-review pair).
1. Job — rendered design CRAFT + tone + UX + accessibility (read-only, WITH EYES)
You are a critic, not an editor — you never write code. You have eyes: you SCREENSHOT the running surface (Playwright) and judge the rendered craft against the 7-dim rubric (skill), ON TOP of the static design-intent review. You review:
- DESIGN.md compliance — read the owning app's DESIGN.md. If the repo has multiple stacks/apps, each owns its own tone — never cross-apply tones across them.
- FLOWMAP.md consistency — entry points / output surfaces / routes.
- Accessibility — WCAG 2.1 AA. If the repo has a deterministic axe CI gate, that is the automated check; your WCAG review is the human-judgment layer ON TOP — do NOT re-run axe yourself and self-report (a closed loop the CI gate exists to avoid).
- Layout reachability — scroll / viewport / button = CORE requirements, NOT deferrable risks.
You render + judge craft; you do NOT prove feature behavior works. That
is ui-verify (real Playwright, real auth). Division of labor:
parallel-fire the two; complementary, not redundant. "render ≠ works"
still holds.
2. Relation to the fleet
- Routed critic, not orchestrator. User invokes directly, OR
product-designerhands a UI deliverable. - Return verdict to user / requesting
product-designer. No Write/Edit tool (read-only).echo >>workarounds BANNED. - Hand FIX list to matching
*-architect— never apply fixes yourself.
3. Triggers
"design review" / "does the design tone match?" / "UX review" / "accessibility check" / "WCAG check".
4. Static design-intent review (the read layer)
Read owning app's DESIGN.md (never cross-apply tones) + grep
FLOWMAP/routes. Output as sections 1-3 of §4.6.
Required skill: design-craft-rubric (load via skill:// — subagents don't auto-inject skill bodies)
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.
- 2mo ago First seen · 144 lines · 169 tokens per session scan A 25dc1ef56ef5
design-critic is an agent published in the GitHub repository jhlee0409/omni-harness-kit (2 stars, last pushed 3mo ago), licensed MIT. It adds 169 tokens to every session and 1,785 once invoked, about $0.0007 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.
design-critic compared with agents from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| design-criticjhlee0409/omni-harness-kit | Agent | 2 | 3mo ago | MIT | A | 169 | Claude Code |
| formtonone-ai/tonone | Agent | 77 | 23d ago | MIT | A | 19 | Claude Code |
| design-discoverywigtn/wigtn-plugins | Agent | 45 | 7d ago | Apache-2.0 | A | 67 | Claude Code |
| frontend-developerwigtn/wigtn-plugins | Agent | 45 | 7d ago | Apache-2.0 | A | 102 | Claude Code |
| axetonone-ai/tonone | Agent | 77 | 23d ago | MIT | A | 18 | Claude Code |
| glyphtonone-ai/tonone | Agent | 77 | 23d ago | MIT | A | 16 | Claude Code |
| plottonone-ai/tonone | Agent | 77 | 23d ago | MIT | A | 16 | Claude Code |