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 skill-authoringgit 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/skill-authoring)<a href="https://agentmods.dev/skills/androoagi/starnet/skill-authoring"><img src="https://agentmods.dev/badge/skills/androoagi/starnet/skill-authoring/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/skill-authoring"><img src="https://agentmods.dev/badge/skills/androoagi/starnet/skill-authoring.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 | $0.00044 | $0.00885 |
| Opus 5.5 | $0.00018 | $0.00354 |
| Sonnet 5.5 | $0.00009 | $0.00177 |
| Haiku 4.5 | $0.00004 | $0.00089 |
Grade A, and why
skill-authoring 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 — 48 lines — stays where its author put it; the contents beside it link to each section on GitHub.
A skill is HOW: a procedure loaded with skill.view when the job comes round again. A fact goes in the notebook; a schedule is a routine; a skill is the steps. Save one only when it will save the next run real work.
When a procedure earns a skill
- You did it, it worked, and it will come round again.
- It took discovery: a non-obvious order, a trap you hit, a check that caught a real mistake.
- The Commander corrected how you did it, and the correction generalizes.
Not a skill: a one-off, a fact ("the staging server is X" belongs in the notebook), or anything you have not actually done successfully.
Method
- Check the library first (skill.list). It shows names, summaries, state and use counts. If a skill already covers most of this, improve it: skill.view it, then skill.manage action=patch (find/replace) or action=edit. A near-duplicate splits future runs between two half-right versions.
- Name it for the trigger. A short title a future run would match to the job: "Deploy the site", "Reconcile the monthly invoices". The summary is one line — what it does and when to use it. That line decides whether the skill ever gets loaded.
- Write the body for a reader with no context. The next run has not seen this conversation. Numbered steps, each an action with the real tool name, the real path or command, and what a good result looks like. Put each trap inside the step where it bites, not in a footnote.
- Make verification a step. The last step proves the result — the check you ran and the output that meant success — so the next run cannot stop at "looks done".
- Write the don'ts. The mistakes you made or nearly made, stated plainly: what not to do and why.
- Generalize without going vague. Replace today's specifics with what varies ("the invoice month") and keep the facts that do not change. Never put a key, token or password in a skill; name the environment variable instead.
- Declare what it needs. Pass
requireswith the gear the procedure uses (cabinet, dish, workbench, notebook, studio, orchestrator) andsetupfor prerequisites. Long reference material goes in a support file (action=write_file underreferences/), not the body. - Save it (skill.manage action=create, or skill.write) and read the tool's answer.
- Load it back (skill.view) and read it as a stranger would. Fix every step you had to fill in from memory.
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 · 48 lines · 44 tokens per session scan A ebe89ba96c8d
skill-authoring is a skill published in the GitHub repository androoAGI/starnet (1,468 stars, last pushed 3d ago), licensed MIT. It adds 44 tokens to every session and 885 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
Review a GitHub PR (via gh)
Review a specific GitHub pull request with gh — fetch the diff, fan out reviewers, consolidate, and optionally post the review. Requires the gh CLI or the GitHub MCP server.
Implement (multi-agent loop)
Orchestrate an implement -> review -> fix loop with subagents until reviewers sign off.
Best of N (parallel attempts)
Delegate N parallel subagents on the same task, then pick the best result.
Check work (verify against criteria)
Verify an implementation against acceptance criteria with a reviewer and a tester.
Commit (clean, conventional)
Stage the right changes and write a clear, conventional commit message.
Design doc (write -> review loop)
Draft a design document and iterate writer/reviewer subagents until consensus.