What problem does it solve? Orchestrating an autonomous coding agent from a terminal requires handling interactive prompts, permission dialogs, session state, and output formats, which is error-prone when done manually. ## Core Features & Use Cases - Two Orchestration Modes: Run one-shot non-interactive tasks with claude -p print mode, or drive multi-turn interactive REPL sessions through tmux with explicit dialog handling for workspace trust and permission prompts. - Structured Output & Session Control: Capture JSON results with cost and token usage, enforce output schemas, stream events, and resume or fork sessions by ID. - Parallel & Automated Workflows: Launch multiple independent Claude instances in tmux, configure hooks, MCP servers, custom subagents, and CLAUDE.md project memory. - Use Case: Pipe a git diff into claude -p to get an automated PR review, or run parallel instances to fix a bug, write tests, and update docs simultaneously. ## Quick Start Ask the agent to run claude -p with your coding task, an allowedTools whitelist, and a max-turns limit in your project directory.