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 agentmods add instructions/starter-series/mcp-server-starter/agents-mdgit clone --depth 1 https://github.com/starter-series/mcp-server-starterWrote 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/instructions/starter-series/mcp-server-starter/agents-md)<a href="https://agentmods.dev/instructions/starter-series/mcp-server-starter/agents-md"><img src="https://agentmods.dev/badge/instructions/starter-series/mcp-server-starter/agents-md.svg" alt="Measured on agentmods" 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 | $0.01497 | $0.01497 |
| Opus 5 | $0.00749 | $0.00749 |
| Sonnet 5 | $0.00299 | $0.00299 |
| Haiku 4.5 | $0.00150 | $0.00150 |
Grade A, and why
mcp-server-starter AGENTS.md 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 4d 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 — 166 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
This file provides context for AI coding assistants (Claude Code, Cursor, Copilot, Cline, etc.)
Project Structure
src/
├── index.ts # Server entry point — registers tools/resources/prompts and starts transport
├── config.ts # Environment variable parsing
├── helpers.ts # Response helpers: ok() and err() (ok supports structuredContent)
├── pkg.ts # JSON-modules import of package.json — sole place {name,version} is read
├── tools/
│ └── greet.ts # Example tool with safety annotations + outputSchema — copy this pattern
├── prompts/
│ ├── hello.ts # Zero-arg prompt example
│ └── code-review.ts # Templated prompt with Zod-validated args
└── resources/
└── server-info.ts # Example resource exposing server metadata
tests/
├── config.test.js # parseConfig environment branches
├── greet.test.js # Tool tests (run against built output in dist/)
├── helpers.test.js # ok() / err() unit tests
├── hello.test.js # Zero-arg prompt tests
├── code-review.test.js # Templated prompt tests
├── server-info.test.js # Resource tests
└── integration/ # End-to-end via SDK in-memory transport + spawned process
├── sdk-contract.test.js
└── server-boot.test.js
Adding a New Tool
- Create
src/tools/your-tool.ts:import { z } from 'zod'; import { ok, err } from '../helpers.js'; export const name = 'your_tool'; export const config = { title: 'Your Tool', description: 'What your tool does', inputSchema: { param: z.string().describe('Parameter description'), }, // outputSchema makes the handler return `structuredContent` validated // against this shape. REQUIRED by the 2026 MCP spec when declared. // Omit for free-form text-only tools. outputSchema: { result: z.string().describe('Processed value'), }, annotations: { readOnlyHint: true, // Does not modify anything destructiveHint: false, // Not destructive idempotentHint: true, // Same input = same result openWorldHint: false, // No external system interaction }, }; export async function handler({ param }: { param: string }) { try { const result = `Result: ${param}`; // Second arg to ok() → structuredContent alongside the text mirror. return ok(result, { result }); } catch (e) { return err(`Failed: ${e instanceof Error ? e.message : String(e)}`); } } - Register it in
src/index.ts:import * as yourTool from './tools/your-tool.js'; server.registerTool(yourTool.name, yourTool.config, yourTool.handler); - Add tests in
tests/your-tool.test.js
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.
- 4d ago First seen · 166 lines · 1,497 tokens per session scan A 637a1a9849e9
mcp-server-starter AGENTS.md is an instructions file published in the GitHub repository starter-series/mcp-server-starter (0 stars, last pushed yesterday), licensed MIT. It adds 1,497 tokens to every session, about $0.0075 per session on Opus 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 2026-08-31.
Other instructions, from other repositories
Repolis AGENTS.md
AGENTS.md instructions for hyeonsangjeon/Repolis, covering ⏱️ 60-second quickstart (no build, no install), 🗺️ repository map (what lives where), ▶️ run & rebuild, 1) run the city (static) and 🚀 deploy.
ai-news-agent AGENTS.md
AGENTS.md instructions for nickzren/ai-news-agent, covering ai news agent, purpose, preferred agent flow, defaults and status artifacts.
ai-news-agent CLAUDE.md
Claude Code instructions for nickzren/ai-news-agent, covering claude notes, short version and important defaults.
EFC-Plugin copilot-instructions.md
Copilot instructions for Nlai741533/EFC-Plugin, covering copilot coding agent instructions, project overview, critical rules, security — never commit secrets and test-first development.
ai-dial-core CLAUDE.md
Claude Code instructions for epam/ai-dial-core, covering claude.md, build & run, set credentials via environment variables, build (skip tests) and run all tests.
blockrun-mcp AGENTS.md
AGENTS.md instructions for BlockRunAI/blockrun-mcp, covering blockrun mcp, commands, project structure, key dependencies and install in codex.