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/stakpak/agent/agents-mdgit clone --depth 1 https://github.com/stakpak/agentWhat 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 | $0.03669 | $0.03669 |
| Opus 5 | $0.01835 | $0.01835 |
| Sonnet 5 | $0.00734 | $0.00734 |
| Haiku 4.5 | $0.00367 | $0.00367 |
Grade A, and why
agent 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 yesterday.
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 — 335 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Stakpak CLI
Agent guidance for working effectively in this codebase.
Project Overview
Stakpak is a security-hardened DevOps AI agent that runs in the terminal. It generates infrastructure code, debugs Kubernetes, configures CI/CD, and automates deployments — without giving the LLM keys to production.
- Language: Rust (edition 2024, nightly features enabled)
- License: Apache-2.0
- Repository: https://github.com/stakpak/agent
Workspace Structure
cli/ # Main binary crate (`stakpak`)
├── src/
│ ├── main.rs
│ ├── commands/
│ │ ├── agent/run/ # Agent execution engine
│ │ │ ├── mode_interactive.rs # Interactive TUI agent loop
│ │ │ ├── mode_async.rs # Async/headless mode
│ │ │ ├── stream.rs # SSE stream processing
│ │ │ ├── checkpoint.rs # Session checkpoint/resume
│ │ │ ├── tooling.rs # Tool execution
│ │ │ └── helpers.rs # Shared helpers
│ │ ├── acp/ # Agent Client Protocol (Zed integration)
│ │ ├── mcp/ # MCP server/proxy commands
│ │ ├── auth/ # Login/account commands (interactive + non-interactive setup)
│ │ ├── autopilot/ # Autopilot: init, up/down, status, schedule, channel
│ │ └── watch/ # Scheduled task runtime (internal, driven by autopilot)
│ ├── config/ # Configuration management
│ │ ├── file.rs # ConfigFile with profiles + ensure_readonly()
│ │ ├── profile.rs # ProfileConfig + readonly_profile()
│ │ └── types.rs # ProviderType (Remote/Local)
│ └── onboarding/ # Interactive setup wizard + save_config.rs
tui/ # TUI crate (ratatui-based)
├── src/
│ ├── app/events.rs # InputEvent / OutputEvent enums
│ └── services/handlers/ # Event handlers (tool, shell, etc.)
libs/
├── ai/ # LLM provider abstraction (`stakai`)
│ └── src/providers/
│ ├── anthropic/ # Anthropic API (convert, stream, types)
│ ├── openai/ # OpenAI-compatible API
│ └── gemini/ # Google Gemini API
├── api/ # API client + local processing (`stakpak-api`)
│ └── src/local/
│ ├── context_managers/ # Message history reduction strategies
│ │ ├── task_board_context_manager.rs # Preserves individual messages
│ │ ├── simple_context_manager.rs # Flattens history to text
│ │ └── file_scratchpad_context_manager.rs
│ └── hooks/ # Context hooks (scratchpad, task board)
├── shared/ # Shared types (`stakpak-shared`)
│ └── src/models/
│ ├── llm.rs # LLMMessage, LLMMessageContent, provider configs
│ ├── stakai_adapter.rs # ChatMessage → StakAI Message conversion
│ └── integrations/
│ ├── openai.rs # ChatMessage, ToolCall, Role types
│ └── mcp.rs # MCP tool call result handling
└── mcp/ # MCP client/server/proxy crates
├── client/
├── server/
└── proxy/
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.
- yesterday First seen · 335 lines · 3,669 tokens per session scan A 376b61864d03
agent AGENTS.md is an instructions file published in the GitHub repository stakpak/agent (1,759 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 3,669 tokens to every session, about $0.0183 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
show-me-the-story AGENTS.md
Instructions for Nigh/show-me-the-story, covering agents.md — ai 小说生成器项目指南, 项目概述, 编译与运行, 完整编译（含前端构建） and 或手动分步.
giselle AGENTS.md
Instructions for giselles-ai/giselle, covering agents.md - giselle development guide, development philosophy, core principle: less is more, guidelines and project overview.
wanwu-workflow CLAUDE.md
Instructions for UnicomAI/wanwu-workflow, covering claude.md, project overview, development commands, environment setup and clone and setup.
giselle CLAUDE.md
Instructions for giselles-ai/giselle, a project described as: Giselle: AI App Builder. Open Source.
intelligent-terminal rust.instructions.md
Concise Rust coding conventions for this repository.
ai AGENTS.md
AGENTS.md instructions for vercel/ai, covering agents.md, project overview, repository structure, key directories and core package dependencies.