Guide

Monitor your Semantic Kernel agents in production.

Semantic Kernel orchestrates plugins and model calls to complete a task. AgentPing instruments the provider and wraps each invocation in a run, so cost, status and output quality are tracked per run.

Track your first AI workflow Jump to setup

Two lines, then every run reports in.

Install the SDK and point it at your Semantic Kernel agent. One record lands per run with the cost, status, latency and output behind it.

shellinstall
pip install agentping
pythonmonitor your agent
import agentping

agentping.init(api_key="apk_…", agent="sk-planner")
agentping.instrument_openai()  # or instrument_anthropic()

with agentping.run(agent="sk-planner") as run:
    result = await kernel.invoke(function, arguments=…)

Semantic Kernel also emits OpenTelemetry; if you already export OTel, point the OTLP exporter at AgentPing instead of wrapping calls. Full reference is in the docs.

What AgentPing tracks for Semantic Kernel.

  • Cost per invocation, across the plugin and model calls.
  • A missed-run alert when a scheduled job stops running.
  • Output quality on the result, scored against your rubric.

See the pillars behind these: Spend, Pulse and Verify.

Can I use OpenTelemetry instead?
Yes. Semantic Kernel emits OpenTelemetry, so if you already export traces you can point the OTLP exporter at AgentPing and skip the wrap. See the OpenTelemetry docs.
Does it work outside Python?
Semantic Kernel also runs on .NET. For non-Python runtimes the OpenTelemetry path is the way in; the agent run is reconstructed from the trace.
Will it slow invocations down?
No. Telemetry runs off the hot path with a hard timeout and a bounded queue.

Monitor your Semantic Kernel agents in minutes.

Add the SDK, send one run, and watch cost, status and quality land in the dashboard.

Track your first AI workflow All monitoring guides