AgentIndex icon
AgentIndex
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Submit Tool
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Home/
Browser Automation/
GenericAgent
GenericAgent logo

GenericAgent

Active·★ 14.3k·MIT·Updated 2026-10-21
★ Most Popular

GenericAgent is a deliberately minimal autonomous agent: about 3K lines of core code, 9 atomic tools and an agent loop of roughly 100 lines. Those tools give an LLM system-level control of the browser, terminal, filesystem, keyboard and mouse, screen vision and Android devices over ADB. The core depends only on requests plus four small packages, with no Playwright or LangChain.

Its design rule is "don't preload skills, evolve them": every newly solved task is crystallized into a Skill, so the agent builds a personal skill tree over time. Advanced capabilities such as web automation, OCR and vision are unlocked by instructing the agent, which installs what is missing and saves the result to memory. The browser tool injects into your real browser, keeping logged-in sessions.

You can use it from a terminal UI, a Streamlit UI or IM bots for Telegram, Discord, Lark, WeChat, QQ and others, with Claude, Gemini, Kimi, MiniMax and other models. It keeps context under 30K tokens. The README warns to use Python 3.11 or 3.12. License is MIT.

This write-up is based on the project's README and public repository data. We have not hand-tested the tool. Data as of 2026-10-22.

License
MIT
Latest release
v0.2.1 · 2026-10-19
Last commit
2026-10-21
Contributors
76
Forks
1,664
Open issues and PRs
200
First published
2026-01-16

Source: GitHub, npm and PyPI public APIs.

#agent-framework#ai-agent#automation#autonomous-agent#browser-automation#claude#computer-control#desktop-automation
$ Install
$ git clone https://github.com/lsdefine/GenericAgent.git && cd GenericAgent && uv venv && uv pip install -e ".[ui]"
↗ Visit site★ GitHub
01

Features

01Self-evolving skill tree: each completed task is saved as a reusable Skill
02TMWebdriver drives your real browser and keeps existing login sessions
03Context kept under 30K tokens, with a core that needs only requests plus four small packages
04Terminal UI, Streamlit UI and IM bots for Telegram, Discord, Lark, WeChat and more
05Works with Claude, Gemini, Kimi, MiniMax and other major models
02

Why choose it

+Skills accumulate from real tasks, so repeated work gets cheaper and more reliable over time.
+Drives your real logged-in browser and Android phone, reaching apps that API-only agents cannot.
+Small, readable core with minimal dependencies and a sub-30K-token context.
03

Trade-offs

!Full system control in your real browser and accounts carries real risk; supervise what it does.
!Many capabilities are set up by the agent itself, which makes environments less predictable.
!Python version constraints (3.11/3.12) and some Windows terminal rendering issues are documented.
04

Maintenance check

ActiveChecked: 2026-10-20
Last commit
2026-10-20
Commits, last 90 days
100+
Releases, last 90 days
2
Latest release
v0.2.1 (2026-10-19)
Open issues
195
Recent issues with at least one comment
8/20
Commits per week, last 12 weeks
Commits per week, last 12 weeks: 14, 11, 4, 16, 4, 4, 4, 2, 5, 0, 15, 11

Oldest week on the left; the latest four weeks are orange. Source: GitHub commit activity.

Status rule: archived on GitHub = Archived; last commit within 90 days = Active; 91 to 365 days = Slowing; older = Stale. All figures come from the GitHub API on the date shown.

Most discussed open issues

  • #16 深圳公务员用上这个了？9 comments · 2026-03-09
  • #345 关于多智能体之间的通信方式，GA自己的分析与回答6 comments · 2026-05-12
  • #813 关于 GenericAgent 的一些修复和问题4 comments · 2026-09-18
  • #814 feishu: card rollover never fires — long tasks silently stop updating3 comments · 2026-09-21
  • #236 feat: 统一GUI与后台架构 — GUI反映并控制后台Agent服务状态3 comments · 2026-04-30

Issue titles are quoted from GitHub. Open the link to check whether a problem still applies to the version you use.

05

Compatibility

Claude / Gemini / Kimi / MiniMax
Models
Verified via docs
Chrome
Real browser
Verified via docs
Android (ADB)
Mobile control
Verified via docs
Telegram / Discord / Lark
IM frontends
Verified via docs
06

Quick start

1
$ git clone https://github.com/lsdefine/GenericAgent.git
2
$ cd GenericAgent
3
$ uv venv
4
$ uv pip install -e ".[ui]"
07

Use cases

↳Automate web tasks in a logged-in real browser
↳Drive Android apps over ADB, e.g. querying expenses in a payment app
↳Personal desktop assistant reachable from chat apps
08

Alternatives

⚡Top VS:
VSGhostDesk★155→VScamofox-mcp★119→VScdpilot★32→VShermes-agent★256.0k→
GhostDesk logo
GhostDesk★ 155
Give any AI agent a full desktop — it sees the screen, clicks, types, and runs apps like a human. Automate anything with a UI: browsers, legacy software, internal tools. No API needed. One Docker command.
VSCompare→
camofox-mcp logo
camofox-mcp★ 119
Anti-detection browser MCP server for AI agents — navigate, interact, and automate the web without getting blocked
VSCompare→
cdpilot logo
cdpilot★ 32
Zero-dependency browser automation CLI. 70+ commands, 10 test assertions, smart commands (click/fill by text — no LLM needed). MCP server for AI agents with 500x fewer tokens. Extract, observe, script runner. 50KB, pure CDP.
VSCompare→
hermes-agent logo
hermes-agent★ 256.0k
The agent that grows with you
VSCompare→
open-claude-in-chrome logo
open-claude-in-chrome★ 255
Claude in Chrome, reverse-engineered and open-source. No domain blocklist. Any Chromium browser. Same 18 MCP tools, same performance.
VSCompare→
safari-mcp logo
safari-mcp★ 223
Native Safari browser automation for AI agents. 80 tools via AppleScript — zero overhead, keeps logins, runs silently in background. Drop-in alternative to Chrome DevTools MCP with 40-60% less CPU/heat on Apple Silicon.
VSCompare→
MCPSafari logo
MCPSafari★ 61
MCPSafari: Native Safari MCP Server for AI Agents
VSCompare→
open-cuak logo
open-cuak★ 392
Reliable Automation Agents at Scale
VSCompare→
See all alternatives →

Comments

Log in to leave a comment

No comments yet. Be the first!

On this page
01Features02Why choose it03Trade-offs04Maintenance check05Compatibility06Quick start07Use cases08Alternatives
lsdefine /
GenericAgent
↗

A ~3K-line agent that controls your computer and turns each solved task into a reusable skill

14.3k1.7kMIT
Last commit1d ago
StatusActive
CategoryBrowser Automation
Contributors76
Latest releasev0.2.1
Links
Documentation↗Discussion↗Issues↗Releases↗
Cloud Infrastructure
Ad

Deploy on DigitalOcean

Host your MCP servers, Docker containers, and AI agents with $200 free credit over 60 days.

Claim $200 Free Credit
© 2026 AgentIndex.app|Built by a 10-year iOS Developer.
Best by use caseAboutContactAdvertisePrivacy PolicyTerms of UseQYSGitHubBuy me a coffee ☕

Browse by Category

MCP ServersCode AssistantWorkflow AutomationRAG / Knowledge BaseMulti-AgentBrowser AutomationLLM InfraDev ToolingObservability

Not affiliated with Anthropic, OpenAI or Microsoft.