Diagnostics¶
Both SDKs stay quiet by default. When something is not working, this is how you make them say why.
Python¶
The SDK logs under the coalex logger and — deliberately — never configures logging for you. A
library that calls basicConfig takes over your application's log configuration, which is not its
to take. So the SDK attaches a NullHandler and leaves the rest to you.
Turn it on:
import logging
logging.basicConfig(level=logging.INFO) # your application's own setup
logging.getLogger("coalex").setLevel(logging.DEBUG)
What that surfaces:
| Logger | What you see |
|---|---|
coalex.context |
which agent id a span was attributed to, and where it came from |
coalex.auto_instrument |
which frameworks were instrumented, and which were found but could not be |
coalex.safe_exporter |
span export failures — the first few, then a heartbeat |
coalex.inventory |
whether the agent's self-declaration was accepted, and the server's reason if not |
Without raising the level you still get warnings and errors, but only through Python's lastResort handler if your application has none of its own — which prints without a timestamp, level or logger name, and cannot be filtered.
TypeScript¶
The SDK writes to console by default. Pass your own logger to route its diagnostics into your
structured logging, or null to silence them entirely:
import { register } from "@coalex-ai/sdk"
register({
apiKey: process.env.COALEX_API_KEY,
logger: {
debug: (m) => myLogger.debug({ sdk: "coalex" }, m),
warn: (m) => myLogger.warn({ sdk: "coalex" }, m),
error: (m) => myLogger.error({ sdk: "coalex" }, m),
},
})
Only warn is required. Every message is prefixed [coalex].
What you will never see in a log line¶
No API key, proxy token or upstream credential reaches a log line or an exception message in either SDK. Raising the level does not change that — it is a property of what is written, not of what is filtered.