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 station-self-checkgit 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/station-self-check)<a href="https://agentmods.dev/skills/androoagi/starnet/station-self-check"><img src="https://agentmods.dev/badge/skills/androoagi/starnet/station-self-check/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/station-self-check"><img src="https://agentmods.dev/badge/skills/androoagi/starnet/station-self-check.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 | $0.00048 | $0.00805 |
| Opus 5.5 | $0.00019 | $0.00322 |
| Sonnet 5.5 | $0.00010 | $0.00161 |
| Haiku 4.5 | $0.00005 | $0.00081 |
Grade A, and why
station-self-check 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 — 37 lines — stays where its author put it; the contents beside it link to each section on GitHub.
A self-check answers "is my station actually working?" from the harness's own readings, never from memory or a guess. Every finding names the tool that showed it and the one action that fixes it.
Method
- Harness first (station.inspect). Build, scheduler (armed, stopped by E-STOP, healthy, last tick error), routine count, connected MCP connectors, recent recorded errors and the last run. A section reported unavailable is itself a finding — call it unknown, not fine.
- Routines (routine.list). For each: enabled, last status, last error, next run. Flag a failed last status; a routine disabled after repeated failures (five in a row pauses it); an enabled routine over a disarmed or stopped scheduler, which will not fire; a next run that is already in the past.
- Loops (loop.list). Stopped, paused, or waiting on a review the Commander has not given — a review loop keeps working until three results wait for review, then pauses until the Commander reviews them.
- Lines and floor (station.layout). Is routing live? One blocking error anywhere stops every line. For each line: its status pill, what blocks it, any paused starts and why, and today's runs, shipped, failed and spend. Bays on no line, and lines with nothing to start them, are idle gear: built, doing nothing.
- Crew work (team.subagents). Your background workers left stale, interrupted or failed.
- Skills (skill.list with
includeArchived: true). Skills marked stale (unloaded for 30 days), archived, used 0 times, or WITHHELD by the skill guard. A skill that never loads usually has a summary that does not match how the work gets asked for, or duplicates another — say which. - Connections (connectors.list, if you have the DISH). What is connected, and anything a routine depends on that is not.
- Rank the findings. Broken and silent first (a routine failing unattended, a blocked line, a stopped scheduler), then waste (idle bays, dead skills), then cosmetic.
- Give one fix per finding, with its owner. You can pause, resume, edit or queue a routine (routine.manage) and patch or archive your own skills (skill.manage), with the Commander's approval where asked. Building, placing props, lifting an E-STOP and forgetting memories are the Commander's.
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 · 37 lines · 48 tokens per session scan A 0d63a39f30f7
station-self-check is a skill published in the GitHub repository androoAGI/starnet (1,468 stars, last pushed 3d ago), licensed MIT. It adds 48 tokens to every session and 805 once invoked, about $0.0002 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
Debug (reproduce -> isolate -> fix)
Systematically diagnose and fix a bug instead of guessing.
html-ppt-zhangzara-block-frame
Rescuing a messy startup deck into a board-grade system — the diagnosis, the page grammar, and the rebuilt proof pages. Built as a decision-grade design craft deck for founders, exec presenter.
build-test
Run the project's build / typecheck / lint / test commands and emit the build.passing + tests.passing signals devloop convergence reads.
patch-edit
Apply one rewrite-plan step at a time as small reviewable file edits, never rewriting whole files when a localised change suffices.
openrig-skills
Use when an agent running inside OpenRig needs the right skill or context for fleet recovery, seat handover, new-seat orientation, a watchdog wake, cross-host reach to an agent on another machine, rig packaging, an OpenRig upgrade, systematic debugging, queue triage, or implementation planning; also when the agent…
oversight-team
Use when you are a seat on the oversight pod (a standing monitor-mode rig that keeps OTHER rigs healthy), configuring or running the drift detectors, or choosing whether to intervene vs escalate. Covers the pull-not-poll posture, the v0 detectors (premature-park, process-drift, off-task, token-burn), the intervention…