FormPilot

FormPilot is a local automation framework that reads structured markdown files and fills out web forms automatically by leveraging Playwright MCP and Azure OpenAI. It runs entirely on your device, with data processing happening locally and Azure OpenAI calls used for intelligent content generation and field interpretation. The system orchestrates a data file (activities-sample.md) through orchestrator.py, which communicates with the browser automation agent to perform form submissions on the target website. Users can benefit from smart field matching, batch processing, retry logic, and optional interactive review before submitting data. The tooling supports configuring processing modes (sequential or batched), adjusting speed with fast mode, and customizing behavior via environment variables and optional custom rules. In practice, you prepare your markdown data, set environment variables for your Azure OpenAI access and target URL, then run the orchestrator to automatically navigate to the site, populate the forms, and manage errors or retries as needed.

Prerequisites:

• Python 3.11 or later
• Node.js (for Playwright MCP, if you plan to use Node tooling alongside)
• Access to Azure OpenAI (Azure OpenAI API key, endpoint, and optional deployment name)

Installation steps:

1. Clone the repository:

git clone https://github.com/yourusername/formpilot-mcp.git
cd formpilot-mcp

2. Create and activate a Python virtual environment (recommended):

python -m venv venv
# On Windows
venv\Scripts\activate
# On macOS/Linux
source venv/bin/activate

3. Install Python dependencies:

pip install -r requirements.txt

4. Create a .env file or export environment variables in your shell with required Azure OpenAI settings and FormPilot options, for example:

export AZURE_OPENAI_API_KEY="your_api_key"
export AZURE_OPENAI_ENDPOINT="https://your-endpoint.openai.azure.com/"
export AZURE_DEPLOYMENT="your_deployment_name"  # optional
export FORMPILOT_FAST_MODE=true
export FORMPILOT_WEBSITE_URL="https://example.com"

5. Prepare your data file (e.g., activities-sample.md) following the documented markdown structure, and ensure orchestrator.py is present in the repository.
6. Run the automation:

python orchestrator.py

7. (Optional) If you plan to customize rules or run in batched mode, modify the configuration in the Python scripts as needed.

Tips and notes:

• Ensure your Azure OpenAI API is properly configured and your deployment supports the required capabilities.
• The framework uses smart field matching to map data fields from Markdown to form controls; if a dropdown cannot be detected, the system falls back to predefined option lists.
• Fast Mode reduces some interactivity for bulk operations but may be less verbose in UI prompts.
• Keep your .env or environment variables secure; do not commit sensitive keys to version control.
• If you encounter failures, review snapshot-based retries and verify that the target website URL in FORMPILOT_WEBSITE_URL is correct and reachable.
• The project emphasizes data locality; adjust privacy-related settings as needed for sensitive data handling.