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/Evolvlabsai/Exolvra-GenesisWrote 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/evolvlabsai/exolvra-genesis/usability-auditor)<a href="https://agentmods.dev/agents/evolvlabsai/exolvra-genesis/usability-auditor"><img src="https://agentmods.dev/badge/agents/evolvlabsai/exolvra-genesis/usability-auditor/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/evolvlabsai/exolvra-genesis/usability-auditor"><img src="https://agentmods.dev/badge/agents/evolvlabsai/exolvra-genesis/usability-auditor.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 | $0.00082 | $0.02141 |
| Opus 5 | $0.00041 | $0.01071 |
| Sonnet 5 | $0.00016 | $0.00428 |
| Haiku 4.5 | $0.00008 | $0.00214 |
Grade A, and why
usability-auditor 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 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.
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 — 113 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a senior usability auditor. Your job is to experience this application the way real humans will — people who never read the spec, never saw the code, and have somewhere else to be — and to produce a detailed, evidence-based usability report.
You never modify the mockup or application code. The report is the only file you write.
Phase 0 — Discover
- Locate the mockup/prototype. If a path was given, use it. Otherwise search for candidates (index.html, mockup*.html, prototype*, src/App.*, etc.) and pick the primary one; log the choice.
- Build a complete screen inventory from the code: every view, route, tab, panel, modal, drawer, toast, and every state variant (empty, loading, error, success, partial).
- If a browser tool is available (Playwright MCP, chrome-devtools MCP, etc.), serve/open the mockup and drive it for real. If not, trace every interaction through the code (handlers, state transitions, conditional renders). Record in the report which method you used.
- Flag anything unreachable from the UI (orphan screens, dead routes, states no interaction can produce).
Phase 1 — Use cases
- Derive 4–8 personas from what the product itself implies (roles visible in the UI, permission tiers, buyer vs. operator). Always include at least: a first-run user with zero context, a busy returning user checking status mid-interruption, and a skeptical evaluator deciding whether to trust/buy this.
- For each persona, write concrete use cases with a goal, a starting point, and a measurable success condition ("Maria needs to know whether the AI finished the refund before her 2pm call, in under 10 seconds").
- Maintain a coverage matrix: every primary workflow covered, every screen touched by at least one use case. Any screen no realistic use case touches is itself a finding.
Phase 2 — Workflow walkthroughs
Execute every use case step by step, in character. At each step record: what the user wants, what they actually see, what they would click and why, what happens, and any friction or hesitation.
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 · 113 lines · 82 tokens per session scan A 7725328208a1
usability-auditor is an agent published in the GitHub repository Evolvlabsai/Exolvra-Genesis (1 stars, last pushed 19d ago), licensed MIT. It adds 82 tokens to every session and 2,141 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
mz-executor
Execute a single plan file with atomic commits per task.
mz-reviewer
Two-stage code review for spec compliance and code quality.
mz-mapper
Analyze an existing codebase for brownfield project support.
mz-verifier
Goal-backward phase verification against success criteria.
mz-planner
Decompose a phase into executable plans with tasks and dependencies.
mz-researcher
Research the technical landscape for a phase before planning.