-
Notifications
You must be signed in to change notification settings - Fork 2.3k
Python: [Feature]: Better Observability for Agent Creation #3624
Copy link
Copy link
Labels
.NETUsage: [Issues, PRs], Target: .NetUsage: [Issues, PRs], Target: .NetobservabilityUsage: [Issues, PRs], Target: observability related featuresUsage: [Issues, PRs], Target: observability related featurespythonUsage: [Issues, PRs], Target: PythonUsage: [Issues, PRs], Target: PythontriageUsage: [Issues], Target: All issues that still need to be triagedUsage: [Issues], Target: All issues that still need to be triaged
Description
Activity
Metadata
Metadata
Assignees
Labels
.NETUsage: [Issues, PRs], Target: .NetUsage: [Issues, PRs], Target: .NetobservabilityUsage: [Issues, PRs], Target: observability related featuresUsage: [Issues, PRs], Target: observability related featurespythonUsage: [Issues, PRs], Target: PythonUsage: [Issues, PRs], Target: PythontriageUsage: [Issues], Target: All issues that still need to be triagedUsage: [Issues], Target: All issues that still need to be triaged
Type
Projects
- StatusShow more project fieldsDone
Description
The create agent span no longer appears to be used. Ideally when we create an agent any noteworthy actions that occur such as MCP initialization should fall under an agent framework span. Here is an example of the MCP calls before any agent framework span. Note the HTTP calls to the MCP server prior to the Invoke Agent span.
Code Sample
Simple working example https://github.com/wdrone/agent_framework_distributed_tracingLanguage/SDK
Both