AppDynamics Python AI Monitoring: Clarification Needed for CircuIT OpenAI-Compatible LLM Support and Missing LLM Metrics

Narendra_Rao — Sun, 22 Mar 2026 11:26:23 GMT

Does AppDynamics Python Agent 25.10.0.8329 support AI auto-instrumentation only for official OpenAI/Bedrock/LangChain SDK traffic, or also our internal cisco circuit OpenAI-compatible endpoint (/openai/... style API)?
If our endpoint is API-compatible but uses non-OpenAI auth/header patterns, is it expected that enable-openai=true will not produce built-in AI Monitoring metrics?
Is there any Python Agent config to map non-OpenAI hostnames to OpenAI instrumentation?
Are there provider allowlists/signature checks in Python GenAI instrumentation that could exclude OpenAI-compatible gateways?
We see successful custom metric registration and reported=True emissions for LLM|..., but metrics are not visible/queryable. Are there controller-side filters/limits/path rules affecting Python custom metrics (INDIVIDUAL + RATE_COUNTER)?
What is the exact expected Metric Browser location and REST metric-path format for Python custom metrics created via appdynamics.agent.api.create_custom_metric(...) in SaaS 25.10?
Is there a known ingestion delay or minimum datapoint threshold before newly created custom metrics appear in UI/API?
If OpenAI-compatible gateways are not natively supported today, what is the recommended supported pattern to monitor LLM calls/tokens/latency in AppDynamics?
Is support for OpenAI-compatible third-party/internal endpoints on the roadmap? If yes, any ETA/version guidance?

Please confirm whether manual custom LLM|... metrics is the correct workaround until native support is available.

topic AppDynamics Python AI Monitoring: Clarification Needed for CircuIT OpenAI-Compatible LLM Support and Missing LLM Metrics in Splunk AppDynamics