Dicklesworthstone/pi_agent_rust

High-performance AI coding agent CLI written in Rust with zero unsafe code

What it solves

pi_agent_rust is a native AI coding agent CLI designed to replace slow, resource-intensive, or unreliable terminal-based AI assistants. It provides a high-performance, single-binary experience with strong security controls for tool execution, reducing the startup overhead and memory footprint associated with managed runtimes like Node.js or Electron.

How it works

Written in Rust, the project is a port of the original Pi Agent. It utilizes two purpose-built libraries: asupersync for structured concurrency and async runtime management, and rich_rust for high-quality terminal UI formatting. The agent supports streaming responses and includes 35 built-in tools. It manages sessions using a SQLite session index and a segmented log for efficient reopening of large histories.

Who it’s for

Developers who want a fast, lightweight AI coding assistant in their terminal that can interact with their codebase, manage long-running sessions, and execute tools securely.

Highlights

  • Native Performance: A single-binary design that minimizes cold-start time and memory usage.
  • Capability-Gated Security: A two-stage execution guard that blocks dangerous shell patterns (e.g., recursive deletes) before they are spawned.
  • Extensive Tooling: 35 built-in tools for coding tasks, with a capability-based hostcall system.
  • Broad Provider Support: Built-in support for local providers like Ollama, llama.cpp, and Mistral.rs, as well as OpenAI-compatible servers.
  • Structured Concurrency: Uses asupersync to ensure predictable resource cleanup and prevent orphaned futures.

Related

  • Project
    ultraworkers/claw-code

    A Rust-based CLI agent harness that allows users to run AI coding agents locally with support for multiple LLM providers.

  • Project
    vinhnx/VTCode

    A secure, open-source terminal coding agent written in Rust that provides a sandboxed runtime and durable sessions for autonomous AI development.

  • Project
    vercel-labs/agent-browser

    A native Rust CLI that lets AI agents (or scripts) control Chrome via simple commands, using accessibility‑based element refs (`@eN`). It supports opening pages, taking snapshots, clicking/filling, extracting text, screenshots, network mocking, tab management, batch execution, and an experimental WebMCP tool system.

  • Project
    aaif-goose/goose

    An open-source, general-purpose AI agent available as a desktop app, CLI, and API that handles coding, research, and automation tasks.

  • Project
    gi-dellav/zerostack

    zerostack is a Rust‑written terminal AI coding assistant that talks to many LLM providers, offers multiple built‑in prompts (code, plan, review, debug, etc.), enforces a granular permission system, supports session persistence, optional sandboxing, and extensible features like advisor models, hooks, and status‑signal sockets—all while staying lightweight (≈ 16 MiB RAM, 26 MB binary).