P

Pi Coding Agent

The coding-agent harness you can make your own. Pi is a minimal terminal coding harness. Adapt Pi to your workflows, not the other way around.

Visit Website
Related by shared tags and keywordsNativePioh-my-openagentkxen
aiai assistantautomation

Overview

Pi is an AI agent toolkit from earendil-works, described as an "AI agent toolkit: unified LLM API, agent loop, TUI, coding agent CLI". The repository is the home of the Pi agent harness project, which includes a self extensible coding agent. The project is distributed as a set of packages, including @earendil-works/pi-coding-agent (an interactive coding agent CLI), @earendil-works/pi-agent-core (an agent runtime with tool calling and state management), and @earendil-works/pi-ai (a unified multi-provider LLM API covering OpenAI, Anthropic, Google, and others). Additional packages listed include pi-tui, pi-durable, pi-telemetry, and chord.

On permissions, the README states that "Pi does not include a built-in permission system for restricting filesystem, process, network, or credential access. By default, it runs with the permissions of the user and process that launched it." For stronger boundaries, the documentation describes three containerization patterns: a Gondolin extension that keeps pi and provider auth on the host while routing built-in tools into a local Linux micro-VM, Plain Docker for running the whole pi process in a local container, and OpenShell for running the whole pi process in a policy-controlled sandbox.

The project is released under the MIT license. The README also invites users of Pi or other coding agents to share their open source sessions, stating that "Public OSS session data helps improve coding agents with real-world tasks, tool use, failures, and fixes instead of toy benchmarks."

Pricing

Detailed plans have not been confirmed in our catalog. Check the official website for current limits and billing terms.

Visit Website

Prices and limits may change. Confirm the currency, billing period, seat minimum and usage caps on the official website.

Use Cases

  • Developers use the interactive coding agent CLI to work with an agent from the terminal.
  • Teams use the unified multi-provider LLM API to access OpenAI, Anthropic, and Google models.
  • Developers use the agent runtime for tool calling and state management.
  • Users containerize or sandbox Pi when they need stronger boundaries.
  • Users run the whole pi process in a local container for simple isolation.
  • Users run the whole pi process in a policy-controlled sandbox.
  • Open source contributors share their coding agent sessions to help improve coding agents.
  • Developers build standalone binaries from the versioned release source archive.

Features

Toolkit for building AI agents

Unified API for LLM access

Built-in agent execution loop

Terminal user interface

Command-line coding agent

Reviews

No reviews yet. Be the first to share your experience!