Powered by Octocode🐙

The missing manual for AI CLI tools.
Deep-dive documentation, implementation secrets, and architecture internals extracted directly from source code.

For developers, researchers, and power users who need to look under the hood. Official documentation tells you what a tool does. We show you how it works, what's undocumented, and how to build on top of it.

Open source doesn't always mean well-documented. Critical details are often buried in code. We extract:

• 🛠️ Internals & Architecture — How the systems are actually built
• 🧩 Undocumented Features — Hidden APIs, beta functionality, and system prompts
• ⚠️ Real-world Gotchas — Edge cases, security models, and implementation constraints
• 🔌 Integration Patterns — How to properly interface with these agents

Anthropic • TypeScript • GitHub Repo

The official SDK for building Claude-powered agents.

• Key Features: 17 built-in tools, MCP integration, specialized sub-agents.
• Deep Dive: Security hooks, permission modes, and agent lifecycle management.

Google • TypeScript • GitHub Repo

Google's terminal CLI for Gemini.

• Key Features: Multi-package architecture, agent-to-agent communication, VS Code integration.
• Deep Dive: Tool definitions, runtime environment, and extension points.

OpenAI • Rust • GitHub Repo

High-performance terminal coding agent.

• Key Features: Rust-based performance, advanced sandboxing, security policies.
• Deep Dive: Execution environment, prompt engineering strategies, and state management.

Community • TypeScript • GitHub Repo

Provider-agnostic AI coding agent for the terminal.

• Key Features: Supports OpenAI/Anthropic/Google/Bedrock, built-in LSP, Client/Server architecture.
• Deep Dive: ACP/MCP protocol integration, TUI implementation, and provider abstraction layers.

@mariozechner • TypeScript • GitHub Repo

Lightweight but powerful CLI coding agent with session management.

• Key Features: Read/Bash/Edit/Write toolset, robust session handling.
• Deep Dive: API reference, data models, and session format specifications.

Found an undocumented tool?
Open an issue with the repo URL. We're always looking for new CLI agents to analyze.

Questions?
Feel free to open discussions about architecture, specific implementation details, or if you need clarification on any docs.

If these docs saved you time or helped you build something cool, please consider starring the repo to help others find it.

Research and educational purposes only. Content derived from public open-source repositories. Verify all information independently. No warranty provided.

Enterprise Support: Need internal documentation for your own AI agents? Reach out: [email protected]