Does it support distributed tracing across services?

Yes, it provides utilities for trace context propagation, allowing you to correlate logs and traces across external service calls using standard headers.

Does this skill support non-Microsoft agent frameworks?

No, this skill is specifically optimized for the Microsoft Agent Framework and its unique observability implementation patterns.

What metrics are automatically tracked by this skill?

It tracks agent invocations, tool call duration, token usage (prompt and completion), error rates, and workflow step execution out of the box.

Can I use this with the .NET Aspire Dashboard?

Yes, it includes specific configurations and Docker setup instructions to integrate your AI agents with the Aspire dashboard for visual debugging.

Can I set up custom alerts for agent failures?

Yes, the skill provides templates for defining SLIs/SLOs and generating Prometheus alerting rules specifically for agent-based latency and error thresholds.

Microsoft Agent Observability

Name: Microsoft Agent Observability
Author: pagerguild

bypagerguild

0•

Analytics & Monitoring

Implements comprehensive monitoring, tracing, and logging for Microsoft Agent Framework applications using OpenTelemetry.

This skill provides expert-level guidance for integrating professional-grade observability into Microsoft Agent Framework systems, enabling developers to monitor performance, trace complex multi-agent workflows, and visualize system health. It simplifies the setup of OpenTelemetry for automatic tracing of LLM calls and tool executions, integrates seamlessly with the .NET Aspire dashboard, and provides standardized patterns for defining service-level objectives (SLOs) and custom alerting rules to ensure production-grade reliability and cost management.

Key Features

010 GitHub stars

02Native support for .NET Aspire Dashboard, Grafana, and Datadog

03Seamless OpenTelemetry integration for traces, metrics, and logs

04Standardized patterns for defining SLIs/SLOs and alerting rules

05Custom metric creation for tracking token usage and execution latency

06Automatic tracing for agent invocations, tool calls, and workflow steps

Use Cases

01Debugging complex agent-to-agent interactions in a multi-agent workflow

02Setting up real-time performance dashboards to identify bottlenecks in tool execution

03Monitoring token consumption and LLM costs across production agent services

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add pagerguild/guilde-lite ms-observability

For use in Claude.ai and ChatGPT

Download Skill