built-in-metrics
Track your AI calls with less instrumentation work.
What it does
Adds LaunchDarkly instrumentation to existing AI calls so users can monitor duration, token usage, success, errors, and streaming latency.
When to use it
Use when an existing codebase needs LaunchDarkly tracking around its AI provider calls.
How to use it
Give it an existing codebase to instrument; it audits the provider call and implements the first tracking tier that fits.
What you provide
- An existing project
Uses
Good to know
- sends a credential to a server: API token
What you need · 6
Python implementations require the LaunchDarkly server-side AI SDK at version 0.20.0 or later; Node implementations use the alternative Node package.
Node implementations require the LaunchDarkly server-side AI SDK at version 0.20.0 or later; Python implementations use the alternative Python package.
Requires an existing AI config from which to obtain a tracker; applications without one must first use configs-create.
Access · 1
This skill
LaunchDarkly
Sends tracking events to LaunchDarkly