What is Agent-S built with?

Agent-S is a Python repo built with FastAPI and uvicorn. It uses AI through openai, PaddlePaddle, together, mcp, scikit-learn, and 2 more, connects to Google, makes its content with beautifulsoup4, imageio-ffmpeg, pillow, pytesseract, and lxml, validates its data with Pydantic, is tested with Selenium, Pytest, and playwright, and deploys with boto3. It has 55,076 written lines, with 80 test files.

Read on November 10, 2026, at commit 0fff0e6. Read it as markdown

Backend

The base 2

AI 7

Services 1

Support

Content 5

Validation 1

Testing 3

  • Pytest49 files
  • playwright1 file
  • Selenium

Deploy 1

Tools 3

  • Black
  • ipykernel1 file
  • tyro1 file

Other 17

  • NumPy14 files
  • pandas
  • python-gitlab
  • pyyaml
  • SciPy
  • websockets
  • requests10 files
  • python-dotenv7 files
  • backoff4 files
  • pyautogui4 files
  • httpx1 file
  • toml1 file
5 more
  • dill
  • nest-asyncio
  • pyobjc
  • pywin32
  • tomli

Other repos with a similar stack

Ordered by how much of its stack they share with Agent-S, then by stars.

Frequently asked questions

What language is Agent-S written in?
Agent-S is written in Python, with 55,076 written lines.
What framework does Agent-S use?
Agent-S is built with FastAPI and uvicorn.
Does Agent-S use AI?
Yes. Agent-S uses AI through openai, PaddlePaddle, together, mcp, scikit-learn, tiktoken, and paddleocr.
What services does Agent-S use?
Agent-S connects to Google.

The proof: gui_agents/s2/core/engine.py, line 1

Does Agent-S have tests?
Yes. Agent-S has 80 test files and is tested with Selenium, Pytest, and playwright.
Which repos have a stack like Agent-S's?

See inside your code, today.

Paste your repo and see it drawn, part by part: a medium one in under a minute.

Free, no account.

Or look inside SeeRepo's own →

Is it your repo? Add its badge, keep it up to date or take it down →