Agent v0: A CLI-first framework to coordinate multiple domain-specific AI agents from one terminal.
A master orchestration tool that breaks down natural language instructions into tasks for specialized AI agents to accomplish concurrently and collates the outputs back to the user.
Security layer developed in Rust: bubblewrap/sandbox-exec for sandboxing processes, AES-256-GCM encryption for keystores (Argon2id key derivation function), SHA-256 hashing + verification for audit logs, and Ed25519 signatures for inter-bot authentication. TypeScript layer takes care of the CLI, daemon, routing on Claude/GPT/Gemini, bots.
In its infancy; latest version (v1.5.0) released recently with M1 support. Questions?
Agent v0: A CLI-first framework to coordinate multiple domain-specific AI agents from one terminal. A master orchestration tool that breaks down natural language instructions into tasks for specialized AI agents to accomplish concurrently and collates the outputs back to the user. Security layer developed in Rust: bubblewrap/sandbox-exec for sandboxing processes, AES-256-GCM encryption for keystores (Argon2id key derivation function), SHA-256 hashing + verification for audit logs, and Ed25519 signatures for inter-bot authentication. TypeScript layer takes care of the CLI, daemon, routing on Claude/GPT/Gemini, bots. In its infancy; latest version (v1.5.0) released recently with M1 support. Questions?