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/eric8810/authy/claude-mdgit clone --depth 1 https://github.com/eric8810/authyWhat 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 | $0.00752 | $0.00752 |
| Opus 5 | $0.00376 | $0.00376 |
| Sonnet 5 | $0.00150 | $0.00150 |
| Haiku 4.5 | $0.00075 | $0.00075 |
Grade A, and why
authy CLAUDE.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 — 59 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md — Project Instructions
What is this
Authy is a CLI secrets store & dispatch tool for AI agents. Built in Rust. Single binary, no server, no accounts.
Build & Test
cargo build # dev build
cargo build --release # release build
cargo test # all tests (unit + integration)
cargo clippy -- -D warnings # lint (must pass clean)
Project Structure
src/cli/— clap command definitions and handlers (one file per command)src/cli/common.rs— shared secret resolution logic (used by run, env, export)src/cli/json_output.rs— JSON response structs for--jsonoutputsrc/cli/env.rs—authy envcommand (output secrets as env vars)src/cli/import.rs—authy importcommand (import from .env files)src/cli/export.rs—authy exportcommand (export as .env or JSON)src/cli/serve.rs—authy serve --mcpcommand (MCP server entry point)src/lib.rs— library crate root, re-exports core modulessrc/api.rs—AuthyClienthigh-level programmatic API facadesrc/mcp/mod.rs— MCP server: JSON-RPC 2.0 handler,McpServerstructsrc/mcp/tools.rs— MCP tool definitions and dispatch (5 tools)src/vault/— encrypted vault storage (age encryption, MessagePack serialization)src/auth/— authentication dispatcher (passphrase / keyfile / session token)src/policy/— glob-based access control policiessrc/session/— HMAC session token generation and validationsrc/audit/— append-only JSONL audit log with HMAC chainsrc/subprocess/— child process spawning with env var injectiontests/api_test.rs— lib-level tests forAuthyClientAPI (serial, isolated HOME)tests/mcp_test.rs— in-memory MCP server tests (JSON-RPC, serial, isolated HOME)tests/integration/— CLI integration tests using assert_cmd + tempfileskills/authy/— Agent Skills standard skill (works with Claude Code, Cursor, OpenClaw, etc.)
Key Conventions
- All secret-holding types must derive
Zeroizeand be zeroized on drop - Secret values flow through stdin/stdout, never CLI arguments
- Diagnostics and errors go to stderr, secret values to stdout
- Session tokens are read-only — no mutation commands accept token auth
- Policy evaluation: deny overrides allow, default deny
- Vault writes use atomic rename (write to .tmp, then rename)
--jsonflag produces structured JSON output on stdout, JSON errors on stderr- Non-interactive mode: fails fast when no TTY and no credentials provided
- Typed exit codes: each error category maps to a specific exit code (1-7)
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 · 59 lines · 752 tokens per session scan A b8ebc064ce89
authy CLAUDE.md is an instructions file published in the GitHub repository eric8810/authy (24 stars, last pushed 6mo ago), licensed MIT. It adds 752 tokens to every session, about $0.0038 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
deer-flow CLAUDE.md
Claude Code instructions for bytedance/deer-flow: The repo's agent guidance lives in AGENTS.md so it is shared across coding agents (Claude Code, Codex, and others). Claude Code imports it below.
zeroclaw CLAUDE.md
Claude Code instructions for zeroclaw-labs/zeroclaw, covering claude.md — zeroclaw (claude code), claude code settings, hooks and slash commands.
nuwax AGENTS.md
Instructions for nuwax-ai/nuwax, covering ai agent system documentation, 系统概述, ai agent 架构, 核心组件 and ai 功能特性.
AgentEval memory-benchmarks.instructions.md
Guidelines for working with AgentEval Memory module — benchmarks, reporting, HTML reports, and LongMemEval integration.
AgentEval copilot-instructions.md
Instructions for AgentEvalHQ/AgentEval, covering agenteval - ai coding agent instructions, architecture overview, environment setup, optional: secondary models for comparison and build & test commands.
framework AGENTS.md
Instructions for ai-driven-dev/framework, covering agents.md, behavior, communication, action and memory management.