automatiq 0.1.0a4

Record browser sessions and reverse-engineer them into scraping scripts.

Your activity, into automation.

AutomatiQ

Alpha (v0.1.0) => Work in progress. Things will break, change, and improve.

Record a browser session, and an AI agent reverse-engineers it into a standalone Python script.

Install

pip install automatiq

Install from source

git clone https://github.com/StoneSteel27/AutomatiQ.git
cd AutomatiQ
pip install -e .

Dev setup

pip install -e ".[dev]"
pre-commit install

This installs ruff, build, twine, and pre-commit hooks (lint + format on every commit).

Configuration

On first run, AutomatiQ creates ~/.automatiq/config.toml with commented defaults. Edit it to override models, timeouts, recording settings, etc.

[models]
agent    = "gemini/gemini-3-flash-preview"
recorder = "gemini/gemini-3.1-flash-lite-preview"
# base_url = "http://localhost:11434/v1"   # Ollama / LM Studio / vLLM

[agent]
max_steps       = 60
sandbox_timeout = 60

[recording]
fps                   = 3
segment_pad           = 2
merge_gap_threshold   = 1.5
max_frames_per_prompt = 8

Priority: CLI flag > ~/.automatiq/config.toml > built-in defaults.

Set your API key in a .env file at the project root (any litellm-supported model works):

GEMINI_API_KEY=your-key-here

Run

# Record a session, then have the agent build a scraper
automatiq run https://example.com

# Or run each step separately
automatiq record https://example.com   # just record
automatiq agent                         # build scraper from last recording

CLI flags override config:

automatiq run https://example.com --model openai/gpt-4o --max-steps 80

What it does

1. Record: Opens Chrome, captures your browsing (video, network requests, user actions).
2. Agent: An LLM investigator reads the session dump, experiments in a sandboxed IPython environment, and produces a working scraping script.

Requirements

• Python 3.11+
• A supported LLM API key (Gemini, OpenAI, OpenRouter, or any OpenAI-compatible endpoint via --base-url)

License

MIT