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 channel-updatesgit 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/channel-updates)<a href="https://agentmods.dev/skills/androoagi/starnet/channel-updates"><img src="https://agentmods.dev/badge/skills/androoagi/starnet/channel-updates/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/channel-updates"><img src="https://agentmods.dev/badge/skills/androoagi/starnet/channel-updates.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 | $0.00036 | $0.00742 |
| Opus 5.5 | $0.00014 | $0.00297 |
| Sonnet 5.5 | $0.00007 | $0.00148 |
| Haiku 4.5 | $0.00004 | $0.00074 |
Grade A, and why
channel-updates 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 — 42 lines — stays where its author put it; the contents beside it link to each section on GitHub.
An update on a phone is read in five seconds between other things. It either tells the Commander something they needed, or it teaches them to mute the station. Most runs should not ping.
When NOT to ping
- Nothing crossed the bar the Commander set. In a routine, reply exactly
[SILENT]— a silent run delivers nothing. - The Commander is talking to you in COMMS right now. Answer there.
- It is progress, not an outcome ("started", "still working") — unless they asked for progress.
- The same news already went out. One change, one message.
- You are not sure it is true. Verify first; a wrong alert costs more than a late one.
Method
- See who you can reach (channel.targets). It lists the chats a person has actually opened with this station, their channel, the agent each is bound to, and whether it is connected right now. You cannot message a new id, number or channel name — reach is widened by a person opening a chat, never by you.
- Lead with the outcome. First line: what happened and what it means — "Build is green on main, ready to ship." Then at most three short lines: the key number, the file, and the one decision needed, if any.
- Fit the phone. One screen. No tables, no headers, no wall of bullets. Plain text is split into platform-sized parts automatically, but a message needing more than eight parts is refused — put the detail in a file.
- Attach instead of pasting. Pass up to four workspace files in
files: images arrive inline, other files as documents. A channel that cannot carry files gets the path named instead. - Ask at most one question, answerable with one word or a number.
- Send (channel.send) and read the result. In APPROVAL mode the Commander confirms the send. A channel that is not connected is refused with nothing sent — report that, and point them to the CHANNELS panel to reconnect.
- For scheduled updates, use the routine's own delivery, not channel.send. An unattended run has nobody to approve a send. Create the routine from the chat that should receive it (routine.create with
deliver: origin) and write the[SILENT]rule into its prompt.
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 · 42 lines · 36 tokens per session scan A 6ca849c2551e
channel-updates is a skill published in the GitHub repository androoAGI/starnet (1,468 stars, last pushed 3d ago), licensed MIT. It adds 36 tokens to every session and 742 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
fs-creative-voltage
OpenDesign's seed pitch: the open, local alternative to closed AI design — why now, the wedge, and the ask. Built as a decision-grade fundraising pitch deck for pre-seed & seed VCs.
ve-midnight-editorial
OpenDesign's financial review: runway, burn, and the sustainability plan that keeps the project independent. Built as a decision-grade data & finance deck for board, leadership.
ib-pitch-book
OpenDesign's investor pitch book: market map, moat, unit economics, and the ask — analyst-grade and diligence-ready. Built as a decision-grade fundraising pitch deck for growth-equity investors.
hps-y2k-chrome
OpenDesign's KPI decision brief: activation and retention drivers, what's working, and the one metric to move next. Built as a decision-grade data & finance deck for CRO, RevOps, leadership.
dcf-valuation
Discounted cash flow valuation and intrinsic value analysis for public companies. Use when the brief asks for DCF, fair value, intrinsic value, price target, undervalued or overvalued analysis, or "what is this company worth?".
html-ppt-zhangzara-bold-poster
OpenDesign's Series A growth story: the traction curve, the expansion motion, and why it's venture-scale. Built as a decision-grade fundraising pitch deck for Series A partners.