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/auth0/auth0-mcp-server/agents-mdgit clone --depth 1 https://github.com/auth0/auth0-mcp-serverWhat 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 | $0.01010 | $0.01010 |
| Opus 5 | $0.00505 | $0.00505 |
| Sonnet 5 | $0.00202 | $0.00202 |
| Haiku 4.5 | $0.00101 | $0.00101 |
Grade A, and why
auth0-mcp-server 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 — 78 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
This file provides guidance to AI coding agents working with this repository.
What This Is
An MCP (Model Context Protocol) server that lets AI assistants interact with the Auth0 Management API. It exposes Auth0 operations as "tools" an LLM can call.
Commands
npm run dev # Run with tsx (no build needed)
npm run build # Compile TypeScript → dist/ (runs format + lint first)
npm run typecheck # Type check without emitting
npm run test # Run vitest once
npx vitest run test/tools/applications.test.ts # Single test file
npm run lint:fix # Auto-fix lint issues
npm run format # Prettier format
Architecture
Request path: src/server.ts validates each tool call (token expiry, rejects undeclared parameters) → src/tools/index.ts dispatches to domain handler → handler calls Auth0 API → src/utils/response-masker.ts redacts secrets before the response reaches the LLM.
Key directories:
src/tools/— one file per domain (applications, actions, logs, forms, etc.). Each exports tool definitions and handlers.src/auth/— two flows:device-auth-flow.ts(browser OAuth2, primary) andclient-credentials-flow.ts(M2M for private cloud).src/clients/— per-client config managers (Claude, Cursor, Windsurf, VS Code, Gemini) that write MCP config into each tool's config file.src/commands/— CLI commands:init(auth),run(start server),logout,session.src/utils/— shared utilities; notable:config.ts(credential loading),tools.ts(filtering logic),response-masker.ts.
Tool System
Tools are named auth0_<verb>_<resource> (e.g. auth0_create_application, auth0_list_logs). Read operations carry readOnly: true in _meta.
To add a tool: define it in src/tools/<domain>.ts, add the handler, register both in src/tools/index.ts.
Tools are filtered at startup by glob patterns (--tools / AUTH0_MCP_TOOLS) and a read-only flag (--read-only / AUTH0_MCP_READ_ONLY=true). Read-only takes priority over patterns.
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 · 78 lines · 1,010 tokens per session scan A f233c1da39d6
auth0-mcp-server AGENTS.md is an instructions file published in the GitHub repository auth0/auth0-mcp-server (115 stars, last pushed 5d ago), licensed MIT. It adds 1,010 tokens to every session, about $0.0051 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 2026-08-30.
Other instructions, from other repositories
authgear-server AGENTS.md
AGENTS.md instructions for authgear/authgear-server, covering agents.md, repository layout, documentation map, common commands and start local dev.
auth0-cli CLAUDE.md
Instructions for auth0/auth0-cli, covering ai agent guidelines for auth0-cli, your role, project overview, project structure and key files.
auth0-cli AGENTS.md
Instructions for auth0/auth0-cli: This repository's guidelines for AI coding agents live in CLAUDE.md, the single source of truth.
deepsecure CLAUDE.md
Instructions for DeepTrail/deepsecure, covering claude.md, project overview, development commands, environment setup and install development dependencies.
authgear-server CLAUDE.md
Claude Code instructions for authgear/authgear-server, a project described as: Open source Auth0/Clerk/Firebase alternative. Passkeys, SSO, MFA, passwordless, biometric login. Self-hosted or cloud. Enterprise-ready for SaaS & mobile apps.
mcp-server-playground AGENTS.md
Instructions for chrisleekr/mcp-server-playground, covering agents.md: ai collaboration guide, project overview, development commands, typescript & code style and mcp implementation patterns.