Plan and inspect
Follow tasks through the orchestrator, tool steps and event stream.
A local workspace for agent tasks
One task. Visible steps.
An alpha Python agent workspace with a terminal, a local web interface, tools, scheduling and an activity record.
Jawa Quant ComputerPROJECT OVERVIEW
01Create a task
02Review tool steps
03Read the result
Illustrated workflow · not a live application session or market feed
01 / Inside the project
Built around the work.
Use local tools
Browser, filesystem and shell tools use the configured permission policy.
Choose a provider
Begin in deterministic demo mode, then configure an OpenAI-compatible or Ollama endpoint.
02 / Explore the workflow
Three steps to understand it.
Plan and inspect
Follow tasks through the orchestrator, tool steps and event stream.
Full setup and implementation ↗03 / Run it yourself
Start from the source.
Python 3.11 or later. Real browser use also needs Playwright Chromium installed.
Clone the repository, then use the commands below. Review the README before connecting external services.
SOURCE CHECKOUT
git clone https://github.com/honeyamn10-source/jawa-quant-computer.git
cd jawa-quant-computer
python -m pip install -e ".[dev]"
jawa-quant-computer
# Or start the local web interface:
uvicorn jqc.api.server:app --host 127.0.0.1 --port 876504 / Know the scope
Clear expectations.
Alpha software. Demo mode is deterministic; a mock result is not proof of a real browser or model operation. Review permissions before enabling tools on personal data.
See GitHub Actions for commit-specific test and build results.