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/grinv/mcp-server-template/agents-mdgit clone --depth 1 https://github.com/Grinv/mcp-server-templateWhat 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 | $0.01296 | $0.01296 |
| Opus 5 | $0.00648 | $0.00648 |
| Sonnet 5 | $0.00259 | $0.00259 |
| Haiku 4.5 | $0.00130 | $0.00130 |
Grade A, and why
mcp-server-template 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 2d 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 — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Single source of truth for working on this repository — for any model or
agent (Claude, and others). CLAUDE.md only references this file; keep all
shared guidance here.
This repository is a template for building TypeScript MCP servers. See TEMPLATE.md for the step-by-step checklist to spin up a new server.
Architecture
A standard stdio MCP server. The design separates a reusable carcass from a thin domain layer, so a new server only rewrites the domain layer.
src/
index.ts # bin entry — calls start()
server.ts # buildServer() + start(); wires clients/tools/prompts [domain wiring]
config.ts # env → validated Config (zod) [domain]
lib/ # GENERIC carcass — reuse as-is:
http.ts # fetch wrapper: timeout, retries, backoff, UA, error mapping
rateLimit.ts # min-interval + sliding-window limiter (e.g. N/s and M/min)
cache.ts # in-memory TTL cache
tokenStore.ts # OAuth refresh-token persistence (use only if your API needs it)
errors.ts # ApiError + status classification + secret redaction
logger.ts # leveled stderr logger (redacts secrets)
result.ts # MCP tool-result builders (jsonResult/errorResult/apiErrorToResult)
clients/ # one client per upstream API, fetch+cache ONLY — no shaping [domain]
format/ # raw upstream payloads → trimmed, agent-facing shapes, one
# *.ts per client plus a co-located *.schemas.ts [domain]
tools/
guard.ts # GENERIC — wraps a tool body, turns errors into results
*.ts # tool registrations [domain]
prompts.ts # optional MCP prompts [domain]
__tests__/ # node:test; helpers.ts is generic infra
scripts/ # build-tests.mjs, run-tests.mjs (generic), check-api.mjs (domain)
Reuse model: copy this template (GitHub "Use this template") and rewrite only
the [domain] parts: config.ts, clients/, format/, domain tools/,
prompts.ts, check-api.mjs, the manifest tool list, and the docs. The lib/
carcass, build tooling, tests infra and CI come for free. Extract lib/ into a
shared package only once duplication across servers actually hurts (YAGNI) —
not before.
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.
- 2d ago First seen · 99 lines · 1,296 tokens per session scan A e2cb62d8d84f
mcp-server-template AGENTS.md is an instructions file published in the GitHub repository Grinv/mcp-server-template (0 stars, last pushed 1mo ago), licensed MIT. It adds 1,296 tokens to every session, about $0.0065 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
tevm CLAUDE.md
Claude Code instructions for evmts/tevm, covering tevm monorepo commands and style guide, project overview, forking implementation, json-rpc support and commands.
moss AGENTS.md
Instructions for D-Robotics/moss, covering agents.md, 仓库身份, 文档所有权与阅读顺序, 代码规范执行摘要（必须遵守） and 第一次进入仓库.
platform AGENTS.md
Instructions for dashpay/platform, covering repository guidelines, project structure & module organization, architecture, technology stack and key components.
agnix typescript.instructions.md
Instructions for agent-sh/agnix, covering typescript instructions and guidelines.
copperhead CLAUDE.md
Instructions for copperheadhq/copperhead, covering claude.md, what this repo is, sources of truth, workflow (openspec) and architecture (per spec.md §2).
openbridge-webcomponents docgen.instructions.md
Instructions for Ocean-Industries-Concept-Lab/openbridge-webcomponents, covering doc generator cli, it never edits the original and keep the prompt aligned with the rules.