ai-agent

pi: Build AI Agents Faster with a Unified LLM Toolkit

pi lets founders create, run, and debug AI agents using a single TypeScript API, interactive TUI, and CLI—ideal for rapid prototyping. Backed by 110k+ GitHub stars.
113,994 stars+2,058★ this week14,489 forksTypeScriptGuide quality 8/10 · reviewed Sep 29Updated 10/9/2026100% free · open source
What it does

Pi provides a unified AI agent toolkit for startup founders, allowing them to leverage a large language model (LLM) API, agent loop, and command-line interface (CLI) tools to automate tasks and workflows.

When to use it
  • •When you need to integrate AI-driven automation into your startup's workflow
  • •When you want to use a unified toolkit for building, testing, and deploying AI-powered agents
  • •When you require a flexible and customizable AI solution with a strong community backing (67k+ GitHub stars)
Ready-to-paste prompt
pi eval 'your_large_language_model_prompts_here' -- to interact with the LLM API and receive a response
Heads up: Make sure you have Node.js (16 or higher) and npm installed on your system before attempting to install and run pi, as it relies on these to function correctly
Saves to your device
Use with Claude
New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

🛠️ Technical setup

Expect 20–40 minutes in a terminal — or let your AI agent drive it.

Try it instantly — no install
Claude Code
mkdir -p ~/.claude/skills/pi && curl -fsSL https://workflowstacks.com/api/skills/pi/claude-skill -o ~/.claude/skills/pi/SKILL.md
Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)
claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

How pi: Build AI Agents Faster with a Unified LLM Toolkit works
Codeflow
Free to inspect

Pi: Build AI Agents Faster with a Unified LLM Toolkit is a very large TypeScript project (~130k lines across 692 code files, plus 486 test files). It is a full software project: use it through its install path rather than reading it end to end. Last commit this month, MIT license, has a test suite.

Size
Very large codebase
~130k lines · 692 code files · days to read — use, don't read
Setup
Developer setup
A real software project. Use it via its install path; don't expect to read it all.
Runs on
Node.js
No API keys detected
TypeScript 94%JavaScript 5%
Where to start reading
  1. 1
    README.md
    Start here — what it does and how to install it
  2. 2
    AGENTS.md
    The instructions the AI actually follows
  3. 3
    package.json
    Dependencies and the commands it exposes
  4. 4
    packages/agent/README.md
    Inside packages/agent/ — an example of what the AI is told to do
What's in each folder
packages/agent/Prompts, skills & agent definitions86 files
packages/server/Backend / API31 files
packages/client/Frontend / UI23 files
packages/coding-agent/Package640 files
packages/ai/Package327 files
packages/tui/Package92 files
scripts/Helper scripts33 files
READMEHas testsDocumentedCI checksMIT licenseUpdated this month
Quick Actions
Details
Creator
earendil-works
Language
TypeScript
Category
ai-agent
Published
8/9/2025

Are you the creator of this tool? and earn 85% of every sale.

Show it off in your README:

Featured on WorkflowStacks
[![Featured on WorkflowStacks](https://workflowstacks.com/api/badge/pi.svg)](https://workflowstacks.com/skills/pi?utm_source=github&utm_medium=badge)
🔥 Hot this week, in your inbox
Skills like this, every Monday.

The five fastest-growing open-source AI skills, ranked by GitHub star growth. One email a week, unsubscribe anytime.