openstat-sdk 0.2.1

OpenStat SDK for AI trading agent telemetry

openstat Python SDK

Python helpers for sending native OpenStat telemetry from AI trading agents.

pip install openstat-sdk

Add editable starter files to an existing Python project:

openstat init

This creates .env.openstat.example, openstat_integration.py, and OPENSTAT.md in the current directory. Existing files are preserved unless you run openstat init --force.

from openstat import OpenStatClient

client = OpenStatClient(
    api_key="ostat_...",
    endpoint="https://api.openstat.online",
    service_name="paper-trader",
    environment="production",
)

run = client.start_agent_run(strategy="breakout")

client.record_run_lifecycle(
    run_id=run["run_id"],
    agent={"id": "agent-1", "name": "Paper Trader"},
    status="running",
    strategy="breakout",
    symbols=["BTC-USD"],
    summary="Run started.",
)

client.record_decision(
    run_id=run["run_id"],
    agent={"id": "agent-1", "name": "Paper Trader"},
    strategy="breakout",
    symbol="BTC-USD",
    venue="paper",
    action="enter_long",
    confidence=82,
    rationale_summary="Momentum and risk budget aligned.",
)

client.record_run_lifecycle(
    run_id=run["run_id"],
    agent={"id": "agent-1", "name": "Paper Trader"},
    status="completed",
    strategy="breakout",
    symbols=["BTC-USD"],
    summary="Run completed.",
)

endpoint defaults to https://api.openstat.online. Set OPENSTAT_ENDPOINT=http://localhost:4000 only when sending telemetry to a local OpenStat API during development.

Use record_run_lifecycle(status="completed") or record_run_lifecycle(status="failed") to settle the matching run on the OpenStat Runs dashboard. Lifecycle events share the same run_id as decisions, risk checks, orders, fills, positions, and PnL snapshots.

Use create_opentelemetry_http_config to get OTLP/HTTP endpoints and headers for traces, logs, and metrics exporters.