ai-agent

pi-agent-desktop

Pi — A cross-platform AI coding agent, bringing the Claude Code experience to your desktop. No environment setup, no terminal commands. Download and start coding right away.
505 stars+19★ this week64 forksTypeScriptUpdated 10/13/2026100% free · open source
What it does

Pi lets you write code by chatting with Claude directly on your desktop—no terminal, no local setup.

When to use it
  • •You need to prototype a feature quickly and want AI‑generated code without configuring a dev environment.
  • •Your team prefers a visual UI over CLI tools for AI‑assisted coding.
  • •You want to generate full files (e.g., React components, API routes) and edit them in place on your machine.
Ready-to-paste prompt
Write a full‑stack Next.js page that provides a JWT‑based login form, validates credentials against a PostgreSQL table `users`, and redirects to `/dashboard` on success.
Heads up: Pi requires a valid Anthropic API key; without it the agent will start but return an "unauthorized" error when you send a prompt.
Saves to your device
Use with Claude
New

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

🔑 Needs an API key

Works after you add credentials — the setup agent will ask you for them.

Try it instantly — no install
Claude Code
mkdir -p ~/.claude/skills/24e38db4-07f4-4469-9814-2e6c90e7939e && curl -fsSL https://workflowstacks.com/api/skills/24e38db4-07f4-4469-9814-2e6c90e7939e/claude-skill -o ~/.claude/skills/24e38db4-07f4-4469-9814-2e6c90e7939e/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-agent-desktop works
Codeflow
Free to inspect

Pi-agent-desktop is a very large TypeScript project (~220k lines across 973 code files, plus 12 test files). Setup is light: installs like a normal app. Reading the code is optional. Last commit this month, MIT license, has a test suite.

Size
Very large codebase
~220k lines · 973 code files · days to read — use, don't read
Setup
One-command install
Installs like a normal app. Reading the code is optional.
Runs on
Node.js
No API keys detected
TypeScript 57%JavaScript 38%CSS 4%Rust 1%
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
    bin/pi-web.js
    Where the program starts running
  4. 4
    package.json
    Dependencies and the commands it exposes
  5. 5
    lib/scheduled-tasks/index.ts
    Inside lib/ — the main logic begins here
What's in each folder
lib/Core code — the actual logic458 files
app/Core code — the actual logic113 files
components/Frontend / UI154 files
bin/Command-line entry points7 files
demo/Examples you can copy263 files
docs/Documentation42 files
hooks/Hooks — scripts that auto-run on events32 files
scripts/Helper scripts28 files
READMEHas testsDocumentedCI checksExamples includedMIT licenseUpdated this month
Quick Actions
Details
Creator
abcwyc
Language
TypeScript
Category
ai-agent
Published
7/22/2026

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/24e38db4-07f4-4469-9814-2e6c90e7939e.svg)](https://workflowstacks.com/skills/24e38db4-07f4-4469-9814-2e6c90e7939e?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.