PromptQL Thread Exporter

Description

A CLI script to fetch thread data from the PromptQL API and output it as JSON, matching the structure in sample_thread_data.json.

Installation

1. Ensure you have Python 3.8 or higher installed.

2. Install dependencies using pip:

   pip install -r requirements.txt
   

3. Create an API key for your PromptQL project. You can create API keys in the PromptQL settings in the playgound

Usage

You can now provide the API key in three ways (highest priority first):

1. CLI flag: --api-key YOUR_API_KEY
2. Environment variable: API_KEY=YOUR_API_KEY
3. Config file: As specified in your ~/.ddn/config.yaml under promptql_secret_keys

Example usage:

python main.py --api-key YOUR_API_KEY --select thread1,thread2

If --api-key is not provided, the script will check the API_KEY environment variable. If neither is set, it will fall back to the config file.

• --project-name: The project name as found in your config.yaml (optional; if not provided, you will be prompted to select a project).
• --project-id: The full project key (e.g., name:env) as found in your config.yaml (optional; takes precedence over project name).
• --select: Optional comma-separated thread IDs to select. If not provided, you will be prompted to select threads interactively. If you press Enter, all threads will be selected.
• --output-dir: Optional directory to save the JSON output files. Defaults to the current directory.

Configuration File:

• The script will read the ddn config file at ~/.ddn/config.yaml containing your PromptQL project secret keys if it exists. See sample_config.yml for an example structure.
• The API key for requests is read from the selected project's secret key in this config file.

Environment Variable:

• BASE_URL: The base URL for the PromptQL API. Set this in your .env file or as an environment variable. Example: BASE_URL=https://promptql.ddn.hasura.app

Private Data Plane:

• Private data planes are supported, but you must specify your BASE_URL to reflect your data plane.

Example

python main.py --project-name alive-urchin-6152 --select 3c0bbb82-82b2-475c-8193-d03e7fee0d42 --output-dir ./output

This will fetch the selected threads and save each to a file named after the thread's title (or thread_id if no title) plus a timestamp, in the output directory.

If you omit --project-name and --project-id, you will be shown a list of projects and can select one interactively.

NOTE: I want to use the PAT stored in ~/.ddn/config.yaml. If anyone knows how to use a PAT to authenticate an API call, let me know.

Error Handling

The script provides user-friendly messages for common HTTP errors:

• 404 (Not Found): "Error: Thread not found with the provided ID."
• 403 (Forbidden): "Error: Unauthorized access. Check your authentication token."
• 500 (Internal Server Error): "Error: Internal server error. Please try again later."

For other HTTP errors, it displays the status code and any error details returned by the server in a formatted manner.

Running Tests

To run the unit tests:

pytest