Skip to content

README.md

Latest commit

 

History

History
104 lines (80 loc) · 3.55 KB

README.md

File metadata and controls

104 lines (80 loc) · 3.55 KB

agentic-layer-sdk-msaf

Microsoft Agent Framework adapter for the Agentic Layer SDK.

This package provides utilities to convert a Microsoft Agent Framework agent into an instrumented A2A Starlette web application.

Usage

import os
from agent_framework import Agent
from agent_framework_openai import OpenAIChatCompletionClient
from agenticlayer.msaf import create_metrics_middleware
from agenticlayer.msaf.agent_to_a2a import to_a2a

agent = Agent(
    client=OpenAIChatCompletionClient(
        model=os.environ.get("AGENT_MODEL", "gemini-2.5-flash"),
        base_url=os.environ.get("LITELLM_PROXY_API_BASE"),
        api_key=os.environ.get("LITELLM_PROXY_API_KEY"),
    ),
    name="MyAgent",
    instructions="You are a helpful assistant.",
    middleware=create_metrics_middleware(),
)
app = to_a2a(agent, name="MyAgent", rpc_url="http://localhost:8000/")
# Then run with: uvicorn module:app

Configuration

Observability

OpenTelemetry setup

Call setup_otel() before creating agents to configure OTLP exporters and enable instrumentation:

from agenticlayer.msaf.otel import setup_otel

setup_otel()

This reads standard OTEL_EXPORTER_OTLP_ENDPOINT / OTEL_EXPORTER_OTLP_PROTOCOL environment variables, sets up trace/log/metric providers, and enables the built-in Agent Framework telemetry layers.

Metrics

The SDK emits the following OpenTelemetry metrics:

Built-in (provided by Agent Framework telemetry layers, enabled by setup_otel()):

Metric Type Description
gen_ai.client.token.usage Histogram Input and output token counts per LLM call
gen_ai.client.operation.duration Histogram Duration of LLM / agent operations

Custom (provided by create_metrics_middleware(), must be added to the agent):

Metric Type Description
agent.invocations Counter Number of agent invocations
agent.llm.calls Counter Number of LLM calls
agent.tool.calls Counter Number of tool calls
agent.errors Counter Number of errors (with error_source attribute)

Add the metrics middleware to your agent:

import os
from agent_framework import Agent
from agent_framework_openai import OpenAIChatCompletionClient
from agenticlayer.msaf import create_metrics_middleware

agent = Agent(
    client=OpenAIChatCompletionClient(
        model=os.environ.get("AGENT_MODEL", "gemini-2.5-flash"),
        base_url=os.environ.get("LITELLM_PROXY_API_BASE"),
        api_key=os.environ.get("LITELLM_PROXY_API_KEY"),
    ),
    instructions="You are a helpful assistant.",
    middleware=create_metrics_middleware(),
)

If you already have other middleware, combine them:

import os
from agent_framework import Agent
from agent_framework_openai import OpenAIChatCompletionClient

agent = Agent(
    client=OpenAIChatCompletionClient(
        model=os.environ.get("AGENT_MODEL", "gemini-2.5-flash"),
        base_url=os.environ.get("LITELLM_PROXY_API_BASE"),
        api_key=os.environ.get("LITELLM_PROXY_API_KEY"),
    ),
    instructions="You are a helpful assistant.",
    middleware=[MyCustomMiddleware(), *create_metrics_middleware()],
)