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 cold-emailgit 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/cold-email)<a href="https://agentmods.dev/skills/androoagi/starnet/cold-email"><img src="https://agentmods.dev/badge/skills/androoagi/starnet/cold-email/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/cold-email"><img src="https://agentmods.dev/badge/skills/androoagi/starnet/cold-email.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 | $0.00047 | $0.02388 |
| Opus 5.5 | $0.00019 | $0.00955 |
| Sonnet 5.5 | $0.00009 | $0.00478 |
| Haiku 4.5 | $0.00005 | $0.00239 |
Grade A, and why
cold-email 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 — 140 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Cold Email
Write business-to-business cold emails and follow-ups that read as if a sharp, thoughtful person wrote them, not a sales machine working through a template. The bar: drafts only, one clear ask per email, and every personal detail traceable to a source you opened in this run.
When to use
- "Write a cold email to X." "Draft outreach for this list." "Nobody replies to my emails."
- A follow-up sequence for prospects who have not answered.
- A rewrite of an outreach email that reads like a pitch.
Not for emails to people who already signed up or bought (email-sequence), and not for finding who to write to (lead-scouting, which comes first).
Method
- Get the brief. Five things: who you are writing to (role, company, why them); the outcome wanted (a reply, an introduction, a call); the value (the specific problem solved for people like them); the proof (a real result, customer or credential, supplied by the Commander); any signals already known. Look for a product brief in the workspace with
fs.searchfirst. Work with what you have. Do not block on a missing input; note what would make the email stronger. - Research, and keep a source table. For individual-level personalisation, read the company's site, news or press page and careers page, and the person's public writing or talks (
web_search, thenweb_fetch;browser.navigateandbrowser.get_textfor pages that need a real browser). Record each usable signal in a table: the detail, the URL, the date on the page, and the exact words you saw. Profile pages behind a login stay closed; use only what the Commander pastes from them. - Choose the personalisation level honestly. Use the four levels below. With no verified signal about the person, write at role or industry level and say that you did. Never dress a guess up as an observation.
- Choose a shape. One of the four common shapes below, or a framework from
references/frameworks.md(included at the end of this skill), or freeform when the email flows without one. - Write the first email. Aim for 25 to 75 words. Follow the principles and the voice notes.
- Write the subject line. Two or three options.
- Write the follow-ups. Three to five emails in total, each with a new angle. Cadence, angles and the last "breakup" email are in
references/follow-ups-and-subject-lines.md(included at the end of this skill). - Run the quality check and fix what fails.
- Deliver drafts and stop. Save the drafts, the source table and the list of gaps with
fs.write, name the file withdeliverable_note, and hand over. The Commander sends.
What ships with it
5 files 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 · 140 lines · 47 tokens per session scan A a2e76de1e808
cold-email is a skill published in the GitHub repository androoAGI/starnet (1,468 stars, last pushed 3d ago), licensed MIT. It adds 47 tokens to every session and 2,388 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.