​
Anthropic Integration

Lucidic supports automatic tracking of Anthropic Claude completions in your agent sessions — with no extra instrumentation needed.

​
How It Works

When you initialize the SDK with: 
lai.init(..., providers=["anthropic"])
Lucidic will:
  • Instrument the Anthropic client using OpenTelemetry
  • Automatically create events for each Claude completion - no manual event creation needed
  • Attach the LLM call to the current active Step (or auto-create one if none exists)
  • Support both sync and async operations

​
What Gets Captured

We automatically capture the following from Anthropic API calls:
  • Input: your messages/prompt to Claude
  • Model: the Claude model used (e.g. claude-3-opus-20240229, claude-3-5-sonnet-20241022)
  • Output: the Claude response (including streaming)
  • Token usage: input and output tokens
  • Cost: calculated based on token usage and model pricing
  • Timing: duration of the API call
  • Images: when using vision capabilities

​
Why This Matters

LLM calls are a core part of most agent workflows — but without visibility, it’s impossible to debug or optimize:
  • Which call caused the failure?
  • Which step was it part of?
  • How much did it cost?
  • What was the actual response?
This integration gives you full transparency, with zero boilerplate.

​
Example

import lucidicai as lai
from anthropic import Anthropic

# Just initialize - that's it!
lai.init(providers=["anthropic"])

client = Anthropic()

# steps and events are automatically created
response = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    messages=[{"role": "user", "content": "Hello Claude!"}]
)

# Session auto-ends when script exits

​
Streaming Example

import lucidicai as lai
from anthropic import Anthropic

lai.init(providers=["anthropic"])

client = Anthropic()

# Streaming responses are also tracked automatically
stream = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1000,
    messages=[{"role": "user", "content": "Explain quantum computing"}],
    stream=True
)

for event in stream:
    if event.type == "content_block_delta":
        print(event.delta.text, end="")

​
Explicit Step Management

import lucidicai as lai
from anthropic import Anthropic

lai.init(
    session_name="claude_task",
    providers=["anthropic"]  # API key and agent ID from env vars
)

client = Anthropic()

lai.create_step(state="Planning", goal="Generate research questions")

# All llm calls are automatically tracked as events within this step
response = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=500,
    messages=[{"role": "user", "content": "What are 5 research ideas about climate change?"}]
)

response = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=500,
    messages=[{"role": "user", "content": "What is the capital of France?"}]
)

lai.end_step()

​
Notes

  • Requires anthropic Python SDK installed (pip install anthropic)
  • All Anthropic client methods are instrumented
  • Both sync and async methods are supported
  • If no step exists when an LLM call is made, Lucidic automatically creates one
  • You can still manually use create_event() for additional context
  • Works with the latest Anthropic Python SDK