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 mcp-server-buildinggit 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/mcp-server-building)<a href="https://agentmods.dev/skills/androoagi/starnet/mcp-server-building"><img src="https://agentmods.dev/badge/skills/androoagi/starnet/mcp-server-building/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/mcp-server-building"><img src="https://agentmods.dev/badge/skills/androoagi/starnet/mcp-server-building.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 | $0.02311 |
| Opus 5.5 | $0.00018 | $0.00924 |
| Sonnet 5.5 | $0.00009 | $0.00462 |
| Haiku 4.5 | $0.00004 | $0.00231 |
Grade A, and why
mcp-server-building 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 — 215 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MCP Server Development Guide
Overview
Create MCP (Model Context Protocol) servers that let AI models work with external services through well-designed tools. The quality of an MCP server is measured by how well it enables a model to accomplish real-world tasks, not by how many endpoints it wraps.
Use this when the Commander wants to integrate an external API or service as an MCP server, in TypeScript (MCP SDK) or Python (FastMCP).
Before you start
- Check the toolchain with
shell.execbefore relying on it:- TypeScript:
node --version(18 or newer) andnpm --version. - Python:
python --version(3.10 or newer) andpip --version. If the one you need is missing, stop and tell the Commander to install it from its official site (nodejs.org or python.org). Do not try to install a language runtime yourself.
- TypeScript:
- Say what will be installed. Building the server installs packages into the project folder (the official MCP SDK, a schema library, an HTTP client). Name them and get the Commander's go-ahead before the first
npm installorpip install. For Python, install into a virtual environment inside the project, not system-wide. - Keys stay out of sight. The server reads the service's key from an environment variable that the Commander sets. Never write a key into code, a committed file, a command, or chat.
- Work in a project folder in your workspace or one the Commander trusted, and write files with
fs.writeandfs.edit.
Process
Building a high-quality MCP server has four phases.
Phase 1: Deep research and planning
1.1 Understand modern MCP design
API coverage and workflow tools. Balance comprehensive coverage of the API's endpoints with specialized workflow tools. Workflow tools are more convenient for specific tasks; comprehensive coverage lets an agent compose operations freely. Which works better varies by client: some benefit from code execution that combines basic tools, others from higher-level workflows. When uncertain, prioritize comprehensive API coverage.
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 · 215 lines · 44 tokens per session scan A 01fe53d56d6c
mcp-server-building 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 2,311 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
agent-operated-software
Use when designing, building, operating, or diagnosing an ongoing application whose live backend or control loop includes OpenRig agents, including applications with a Markdown, YAML, or JSON agent control plane or a thin surface over specialist agent roles.
oma-architecture
Evaluate system boundaries and architectural tradeoffs. Use for architecture decisions, design reviews, and ADRs.
service-app-creator
Create or update the Service component of a NextClaw Mini App, choosing between Portable Rust/WASI Components and native-process MCP services. Use after nextclaw-app-creator selects a Service-backed app, or when the user explicitly asks for Service Actions, portable components, local files, external APIs, commands…
architecture
Greenfield architecture design: map functionality flows, draw components, design APIs, classify dependencies, plan observability. For multi-component, API, schema, auth, or integration-heavy work. For retrofitting existing code, use codebase-hygiene instead.
yao-process
Yao process execution expert. ALWAYS invoke this skill when the user needs to call a Yao process, query data models, run scripts, or check process permissions. Do not call processes without checking this skill first.
byok-relay
OpenAI-compatible LLM gateway for any client-side application (browser, mobile, React Native, Flutter, VS Code extensions, browser extensions, Electron, smart TV, and more). Routes requests to OpenAI, Anthropic, Gemini, Groq, Mistral, and 200+ models, handling CORS, key encryption, and streaming without a dedicated…