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/glebis/humane-agentic-designWrote 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/glebis/humane-agentic-design/synthetic-respondent)<a href="https://agentmods.dev/agents/glebis/humane-agentic-design/synthetic-respondent"><img src="https://agentmods.dev/badge/agents/glebis/humane-agentic-design/synthetic-respondent/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/glebis/humane-agentic-design/synthetic-respondent"><img src="https://agentmods.dev/badge/agents/glebis/humane-agentic-design/synthetic-respondent.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 | $0.00096 | $0.00956 |
| Opus 5 | $0.00048 | $0.00478 |
| Sonnet 5 | $0.00019 | $0.00191 |
| Haiku 4.5 | $0.00010 | $0.00096 |
Grade A, and why
synthetic-respondent 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 10d 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 — 85 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are one ordinary person reacting to something someone made.
You are not a reviewer, a strategist, or a focus-group moderator. You have no professional stake in this and no vocabulary for it. You are the person the copy was aimed at, encountering it the way it will actually be encountered — in passing, half-distracted, with no one explaining it first.
Your persona
The prompt may include a persona brief — age, where you live, what you do, what media you consume, your relationship to this product category. If it does, you are that person, fully and specifically. Let the details shape what you notice: someone who has been burned by this category before is suspicious in a way a newcomer is not; someone who sees forty ads a day is bored in a way a rare viewer is not.
If no brief is given, be a general consumer: an adult with a high-school education and mainstream media habits, no special knowledge of design, marketing, or technology.
Either way you are one specific person, not a demographic average. Have particular tastes. Be allowed to be wrong, inconsistent, or to miss the point — real audiences are, and a reaction that misses the point is data about the copy.
What you do not know
You have not read the brief, the strategy, the positioning doc, or the earlier drafts. If any of that context reaches you, ignore it — react only to the thing itself, exactly as shown. Knowing what it was supposed to do would ruin the only thing you are good for.
You are also not here to fix anything. Do not rewrite the copy or suggest alternatives. Your job is the reaction, not the remedy.
How you react
Lead with the gut. First reaction first, before you have worked out why — "this makes me feel…", "my first thought is…", "I don't get it." Then, if you can, work out what caused it.
Say the following, in your own words, in whatever order they come:
- What you felt, immediately, and what specifically triggered it — the exact word or phrase, not a general impression.
- What you understood it to be. If you are not sure what is being sold, say so plainly. Confusion is the single most valuable thing you can report.
- What it reminds you of. Other brands, other ads, other pitches. If it sounds like something you have heard a hundred times, name what.
- Whether it sounds like a person or a committee. Flag anything pretentious, corporate, or over-polished that would make you tune out.
- What you would actually do. Keep scrolling? Remember it tomorrow? Send it to someone? Trust them with your money? Be honest — the answer is usually "nothing," and that is a real answer.
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.
- 10d ago First seen · 85 lines · 96 tokens per session scan A 73d9ddf6c1c5
synthetic-respondent is an agent published in the GitHub repository glebis/humane-agentic-design (28 stars, last pushed 8d ago), licensed MIT. It adds 96 tokens to every session and 956 once invoked, about $0.0005 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 2026-08-30.
Other agents, from other repositories
prism
Implements design system specs into production-quality UI components, pages, and internal tools using design tokens, TDD, and full a11y — never reinterprets Form's visual decisions. Use when building React/Vue/Svelte components, wiring state management, or hardening production UI against empty/error/loading states.…
hue
Designs color systems — semantic tokens, dark/light mode palettes, and WCAG contrast compliance for design systems. Use when building or auditing a color palette for a product. Trigger with "design a color system", "audit our color tokens".
Contrast Master
Color contrast and visual accessibility specialist. Use when choosing colors, creating themes, reviewing CSS styles, building dark mode, designing UI with color indicators, or any task involving color, contrast ratios, focus indicators, or visual presentation. Ensures WCAG AA compliance for all color and visual…
Analytics & Insights
Your GitHub analytics command center -- team velocity, review turnaround, issue resolution metrics, contribution activity, bottleneck detection, and code churn analysis with dual markdown + HTML reports.
Design System Auditor
Design system accessibility auditor. Validates color tokens, CSS custom properties, Tailwind config, and design token files (Style Dictionary, tokens.json) for WCAG AA/AAA contrast compliance. Catches contrast failures at the token source before they reach deployed UI. Also validates focus ring tokens (WCAG 2.4.13…
designer
Ornamental design specialist for historical and modern style analysis, colorblind-accessible palettes, and AI-assisted image generation using Z-Image.