Skip to main content

​
Install

pip install trulayer anthropic

​
Instrument

import os
from anthropic import Anthropic
import trulayer

trulayer.init(api_key=os.environ["TRULAYER_API_KEY"], project_name="my-app")

client = Anthropic()
trulayer.instrument_anthropic(client)

​
What gets captured

Every call to messages.create (sync and streaming, async and sync) becomes an llm span with:
  • input — the full messages array and system prompt
  • output — the response content blocks (text + tool_use)
  • model
  • prompt_tokens → Anthropic’s input_tokens
  • completion_tokens → Anthropic’s output_tokens
  • cache_creation_input_tokens, cache_read_input_tokens → attached as metadata when prompt caching is in use
  • stop_reason — attached as metadata (end_turn, tool_use, max_tokens)
  • latency_ms

​
Streaming

with client.messages.stream(
    model="claude-sonnet-4-6",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hi"}],
) as stream:
    for text in stream.text_stream:
        print(text, end="")
# span closes when the `with` block exits with final token counts

​
Tool use

Claude’s tool calls appear in the span output under content[].tool_use. To trace tool execution as its own span: 
with trulayer.current_trace().span("tool:get_weather", span_type="tool") as span:
    span.set_input(tool_input)
    result = get_weather(**tool_input)
    span.set_output(result)
Then feed the result back to Claude with a follow-up messages.create call — that becomes a second llm span, and the two-turn conversation appears as a clear waterfall in the dashboard.

​
Prompt caching

Anthropic’s prompt caching works transparently — the SDK records cache_creation_input_tokens and cache_read_input_tokens as span metadata so you can see cache hit rates in the dashboard’s Metrics view. Filter by metadata.cache_hit_rate or query via the API: 
curl "https://api.trulayer.ai/v1/metrics?project_id=my-app&from=...&filters[model]=claude-sonnet-4-6" \
  -H "Authorization: Bearer $TRULAYER_API_KEY"

​
Disabling

trulayer.uninstrument_anthropic(client)

​
Known gotchas

  • Extended thinking — thinking content is captured as part of the span output (content[].thinking). It’ll show up in trace detail views alongside the visible response.
  • Batch API (messages.batches) — not auto-instrumented. Wrap batch submissions manually with trace() if you want observability.
  • Vertex and Bedrock — use the respective Anthropic SDK client (AnthropicVertex, AnthropicBedrock) and the same instrument_anthropic() call.