HeyAgent is a local-first desktop AI agent that can see, understand, and operate your computer via CLI, desktop app, or Telegram, with real desktop control, browser automation, Google Workspace integration, and optional voice replies.

Stars

5

7-day growth

No data

Forks

0

Open issues

0

License

AGPL-3.0

Last updated

2026-07-30

AI repository intelligence
FR-AI / ANALYSIS

Why it is worth attention

It goes beyond chat by directly controlling the user's existing desktop and real browser profile, supports remote control through Telegram, and includes deterministic orchestration with verification, model failover, and a policy-based permission system.

Who it is for

  • Developers and power users who want an open-source local desktop AI agent
  • Russian-speaking users who prefer Russian-first, typo-tolerant commands
  • People who need to control their Windows PC remotely via Telegram
  • Users who want to automate Google Workspace and desktop workflows

Use cases

  • Automating desktop tasks like opening apps, controlling volume, and typing in notepad
  • Using a real browser profile to search, fill forms, or complete web tasks
  • Creating and reading Google Docs, Sheets, Slides, Drive files, and Calendar events via natural language
  • Sending Telegram messages to contacts through Telegram Desktop and continuing conversations until a configured condition

Strengths

  • Local-first design stores credentials and state under ~/.heyagent and keeps OAuth tokens outside the repository
  • Real desktop and browser control across Windows, macOS, and Linux, with Windows as the best-covered platform
  • Supports multiple model providers including OpenAI, AWS Bedrock, OpenRouter-compatible services, and local Ollama
  • Includes policy modes, Telegram chat pairing, loopback-only gateway defaults, and approval prompts for sensitive actions

Considerations

  • Windows 10/11 is recommended and the most thoroughly tested platform, while macOS and Linux have partial system controls
  • Google Workspace setup requires manually creating a Google Cloud project and enabling five separate APIs, with Testing-mode authorizations expiring after seven days
  • Broad permissions can be risky; users must understand the security model and approval policies before enabling full access

README quick start

Quick start

Description

Your Own Autonomous Computer Agent

Related repositories

Similar projects matched by category, topics, and programming language.

makecindy
Featured
makecindy GitHub avatar

cindy

Cindy is an open-source AI agent that runs locally on your machine, integrates multiple AI harnesses and models, and provides memory, skills, and automation to perform real work in your projects and apps.

AI & Machine LearningLarge Language Models
958
hahhforest
Featured
hahhforest GitHub avatar

pi-textbook

A hands-on course that walks through building a Pi-style coding agent from scratch across 15 checkpoints, starting from an offline agent trajectory.

AI & Machine LearningLarge Language Models
625
deerwork-ai
Featured
deerwork-ai GitHub avatar

deer-workflow

An open-source Dynamic Workflow runtime that combines deterministic TypeScript orchestration with replaceable Agent runtimes.

AI & Machine LearningLarge Language Models
312