StarNet is a local-first desktop harness for creating AI agents and arranging them in a pixel-art space station whose layout represents their capabilities, permissions, and workflow. It lets people run real model calls and tools through multiple bounded agents, using their own provider keys or accounts and optionally messaging the station through chat services. The catalogue entries are skills that extend or support work performed by StarNet agents.
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 androoAGI/starnet --skill 1.0.0git clone --depth 1 https://github.com/androoAGI/starnetWrote 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/androoagi/starnet/1.0.0)<a href="https://agentmods.dev/skills/androoagi/starnet/1.0.0"><img src="https://agentmods.dev/badge/skills/androoagi/starnet/1.0.0/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/androoagi/starnet/1.0.0"><img src="https://agentmods.dev/badge/skills/androoagi/starnet/1.0.0.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 | $0.00031 | $0.00549 |
| Opus 5.5 | $0.00012 | $0.00220 |
| Sonnet 5.5 | $0.00006 | $0.00110 |
| Haiku 4.5 | $0.00003 | $0.00055 |
Grade A, and why
accessibility-audit 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 12d 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 — 33 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Accessibility work goes wrong when it becomes a list of automated warnings nobody acts on. The useful version checks the small number of failures that genuinely prevent someone from using the page, and proves each one on the live page.
Method
- Walk the page with the keyboard only. Tab through every interactive element in order. Anything reachable by mouse but not by keyboard is a hard blocker, as is a focus trap you cannot escape, or a tab order that jumps around the page.
- Check that focus is VISIBLE. A removed focus outline with nothing in its place makes keyboard use impossible even when the tab order is correct.
- Check names on controls. Every button, link, input, and icon-only control needs an accessible name. Read the page's structure and find controls whose name is empty, or is "button", or is a filename.
- Check contrast on real rendered colours — body text, muted/secondary text, placeholder text, text on images, and disabled states, which are the ones that usually fail.
- Check structure: one main heading, headings that descend without skipping, form inputs tied to labels, images with alt text that says what the image MEANS (and empty alt for decoration).
- Check motion and state: anything that animates or auto-plays, and whether errors are announced rather than only shown in red.
- Rank by who is locked out. A keyboard trap outranks a heading-order warning by a wide margin.
Rules
- Test the live rendered page, not the source — computed colours and injected markup are what users actually meet.
- Every finding names the element and gives the concrete fix, not a WCAG number on its own.
- Never claim compliance. Automated and manual checks find barriers; they do not certify a standard. Say what was checked and what was not.
- Real assistive-technology testing by a human is the finish line — say so when the stakes warrant it.
Output
Blockers first (keyboard, focus, unnamed controls), then serious issues, then polish — each with the element, what it breaks for whom, and the fix. Plus what you could not check.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 12d ago First seen · 33 lines · 31 tokens per session scan A 468b148e3ede
accessibility-audit is a skill published in the GitHub repository androoAGI/starnet (1,468 stars, last pushed 3d ago), licensed MIT. It adds 31 tokens to every session and 549 once invoked, about $0.0001 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 2026-10-04.
Other skills, from other repositories
web-prototype
General-purpose desktop web prototype. Single self-contained HTML file built by copying the seed assets/template.html and pasting section layouts from references/layouts.md. Default for any landing / marketing / docs / SaaS page when no more specific skill matches.
velar-luxury-real-estate
Use this plugin when the user wants a high-end luxury real-estate / architecture landing page with cinematic scroll choreography: a typewriter preloader that lifts away, a scroll-driven house image that rises from below and scales up while pinning to a dark statement section, a sticky dark stats band with count-up…
web-prototype-taste-soft
Apple-tier soft web prototype. Silver/cream canvas, double-bezel cards, button-in-button CTAs, generous squircle radii, spring motion, ambient mesh. Distilled from Leonxlnx/taste-skill soft-skill + sections 4–8 of taste-skill.
mockup-device-3d
Static iPhone and MacBook 3D-style showcase with real HTML embedded on screens, glass-lens refraction, and 360-degree turntable composition.
web-prototype-taste-brutalist
Swiss industrial-print web prototype. Newsprint canvas, monolithic black grotesque, viewport-bleeding numerals, hairline grid dividers, hazard-red accent, ASCII syntax decoration. Distilled from Leonxlnx/taste-skill brutalist-skill (Swiss Industrial Print mode).
web-prototype-taste-editorial
Editorial-minimalist web prototype. Warm monochrome canvas, serif display + grotesque body, 1px hairline borders, muted pastel chips, generous macro-whitespace, ambient micro-motion. Distilled from Leonxlnx/taste-skill minimalist-skill.