loopengine 1.1.0

Official LoopEngine SDK for sending feedback to the Ingest API

LoopEngine

Official LoopEngine SDK for sending feedback to the Ingest API. Two-line usage: create a client with your credentials, then call send with your payload.

• No external dependencies — uses the Python standard library for HTTP and crypto
• Small surface — one main client (LoopEngine) plus an async wrapper (AsyncLoopEngine)

Install

pip install loopengine

Usage (sync)

from loopengine import LoopEngine

client = LoopEngine(
    project_key="pk_live_...",
    project_secret="psk_live_...",
    project_id="proj_...",
)

result = client.send({"message": "User reported a bug"})
if result.ok:
    print(result.body)  # e.g. {"id": "fb_...", "analysis_status": "pending"}

Usage (async)

import asyncio
from loopengine import AsyncLoopEngine


async def main() -> None:
    client = AsyncLoopEngine(
        project_key="pk_live_...",
        project_secret="psk_live_...",
        project_id="proj_...",
    )
    result = await client.send({"message": "User reported a bug"})
    if result.ok:
        print(result.body)


if __name__ == "__main__":
    asyncio.run(main())

Config

Obtain project_key, project_secret, and project_id from your LoopEngine dashboard. A typical configuration pattern is to read them from environment variables:

import os
from loopengine import LoopEngine

client = LoopEngine(
    project_key=os.environ["LOOPENGINE_PROJECT_KEY"],
    project_secret=os.environ["LOOPENGINE_PROJECT_SECRET"],
    project_id=os.environ["LOOPENGINE_PROJECT_ID"],
)

Payload

The payload object you send must match the fields and constraints you defined when creating your project in the LoopEngine dashboard (required fields, allowed keys, value types, etc.). At a minimum, it should include all the required fields according to your project's schema.

You do not need to pass project_id in the payload; it is automatically injected from the client configuration.

Payloads can be:

• A mapping/dict (dict[str, object])
• Any JSON-serializable object (for example a dataclass) that encodes to a JSON object

Geolocation

You can send device location so feedback is associated with coordinates instead of IP-based geo. Pass optional keyword-only arguments geo_lat and geo_lon to send(). When both are provided, the SDK adds geo_lat and geo_lon to the request body; they are included in the HMAC signature. Omit both to use IP-based geolocation. The API expects valid ranges: latitude -90 to 90, longitude -180 to 180.

# Without geo (IP-based location is used)
result = client.send({"message": "Feedback"})

# With device coordinates
result = client.send(
    {"message": "Bug at my location"},
    geo_lat=34.05,
    geo_lon=-118.25,
)

Quick test with uv and clienttest

This repository includes a small clienttest example app you can run to verify your credentials and connectivity.

1. Install uv (a fast Python package manager/runner):

   pip install uv
   # or: pipx install uv
   

2. From the loopengine-python directory, run the example:

   # Using environment variables (recommended)
   export LOOPENGINE_PROJECT_KEY="pk_live_..."
   export LOOPENGINE_PROJECT_SECRET="psk_live_..."
   export LOOPENGINE_PROJECT_ID="proj_..."
   
   uv run examples/clienttest.py
   

   uv run creates an isolated environment, resolves dependencies, and executes the script in one step. Because this SDK has no runtime dependencies beyond the standard library, uv mainly provides a fast, reproducible way to run the example without managing a separate virtualenv.

3. Alternatively, edit placeholders directly in examples/clienttest.py:

   project_key = "<your_project_key_here>"
   project_secret = "<your_project_secret_here>"
   project_id = "<your_project_id_here>"
   

   Then run:

   uv run examples/clienttest.py
   

Development

To run tests locally:

uv run pytest

Requirements

• Python >= 3.9

License

MIT