HoneyHive Docs

With HoneyHive, we allow users to get visibility into their AI applications by tracing:

Model inference calls as model events
External API calls (like retrieval) as tool events
Collections of workflow steps as chain events
An entire trace of requests as a session, which includes back-and-forth user turns

Logging a Trace

We use OpenTelemetry to automatically instrument your AI application. You can find the full list of supported packages here.

Prerequisites

You have already created a project in HoneyHive, as explained here.
You have an API key for your project, as explained here.

Expected Time: 5 minutes

Installation

To install our SDKs, run the following commands in the shell.

pip install honeyhive

npm install honeyhive

pip install honeyhive-logger

npm install @honeyhive/logger

Authenticate the SDK & initialize the tracer

Initializing the HoneyHiveTracer marks the beginning of your session and allows you to begin tracing your program. To initialize, pass the following parameters:

api_key: Get your API key from HoneyHive Account Settings.
project: Create a project from HoneyHive Projects.
(optional) source: An environment variable for the trace, e.g. “prod”, “dev”, etc. Defaults to “dev”.
(optional) session_name: A custom name for your agent session. Defaults to the main module name.

If you are using a self-hosted or dedicated deployment, you also need to pass:

server_url: The private HoneyHive endpoint found in the Settings page in the HoneyHive app.

from honeyhive import HoneyHiveTracer

# Add this code at the beginning of your AI pipeline code
HoneyHiveTracer.init(
    api_key=MY_HONEYHIVE_API_KEY,
    project=MY_HONEYHIVE_PROJECT_NAME,
    source=MY_SOURCE, # Optional
    session_name=MY_SESSION_NAME, # Optional
    server_url=MY_HONEYHIVE_SERVER_URL # Optional / Required for self-hosted or dedicated deployments
)

# Your LLM and vector database calls will now be automatically instrumented
# Run HoneyHiveTracer.init() again to end the current session and start a new one

import { HoneyHiveTracer } from "honeyhive";

// Add this code at the beginning of your AI pipeline code
const tracer = await HoneyHiveTracer.init({
  apiKey: MY_HONEYHIVE_API_KEY,
  project: MY_HONEYHIVE_PROJECT_NAME,
  source: MY_SOURCE, // Optional
  sessionName: MY_SESSION_NAME, // Optional
  serverUrl: MY_HONEYHIVE_SERVER_URL // Optional / Required for self-hosted or dedicated deployments
});

// Make sure to await the trace call when using async functions
await tracer.trace(async () => {
  // Your AI pipeline code here
  
  // Note: Auto-instrumentation is only supported for CommonJS implementations
  // Note: For ESModules implementations, please refer to Step 3 below

  // Your async AI pipeline code here
  const result = await someAsyncFunction();
  // ... more async code ...

});

// Instantiate a new tracer object with HoneyHiveTracer.init() to trace a new session

from honeyhive.utils.langchain_tracer import HoneyHiveLangChainTracer

honeyhive_tracer = HoneyHiveLangChainTracer(
    api_key=MY_HONEYHIVE_API_KEY,
    project=MY_HONEYHIVE_PROJECT_NAME,
    name=MY_SESSION_NAME,
    source=MY_SOURCE, # e.g. "prod", "dev", etc.
    metadata=MY_METADATA, # optional field
    base_url=MY_HONEYHIVE_SERVER_URL # optional / required for self-hosted or dedicated deployments
)

# Your LangChain code goes here

# When invoking your LangChain agent, chain, tool, or retriever, add the tracer to the callbacks
agent_result = agent(
    "Which city is closest to London as the crow flies, Berlin or Munich?",
    callbacks=[honeyhive_tracer],  # Add the tracer to the callbacks
)

# The agent execution will now be traced and logged to HoneyHive

import { HoneyHiveLangChainTracer } from "honeyhive";

async function main() {
    const config = {
        apiKey: MY_HONEYHIVE_API_KEY,  // Your HoneyHive API key for authentication
        project: MY_HONEYHIVE_PROJECT_NAME,  // The name of your HoneyHive project
        source: "dev",  // The source of the trace, e.g., "dev", "prod", etc.
        sessionName: "Langchain JS Quickstart",  // A name for this tracing session
        baseUrl: MY_HONEYHIVE_SERVER_URL // optional / required for self-hosted or dedicated deployments
    });

    // Create a new LangChain tracer instance
    const tracer = new HoneyHiveLangChainTracer(config);

    // Start a new tracing session
    await tracer.startNewSession();

    // Your LangChain JS code goes here

    // When invoking your LangChain agent, chain, tool, or retriever, add the tracer to the callbacks
    await agentExecutor.invoke(
        { input: "What is task decomposition?" },
        { callbacks: [tracer] }  // Add the tracer to the callbacks
    );

    // The agent execution will now be traced and logged to HoneyHive
}

main().catch(console.error);

from honeyhive import HoneyHiveTracer

# add this code at the start of your LlamaIndex script
HoneyHiveTracer.init(
    api_key=MY_HONEYHIVE_API_KEY,
    project=MY_HONEYHIVE_PROJECT_NAME,
    server_url=MY_HONEYHIVE_SERVER_URL # optional / required for self-hosted or dedicated deployments
)

# Your LlamaIndex session will now be automatically instrumented
# Run HoneyHiveTracer.init() again to end the current session and start a new one

# The honeyhive-logger package provides a lightweight, stateless,
# dependency-free way to send session and event data to HoneyHive.
from honeyhive_logger import start

session_id = start(
    api_key=MY_HONEYHIVE_API_KEY,
    project=MY_HONEYHIVE_PROJECT_NAME,
    session_name="v1",
)

// The honeyhive/logger package provides a lightweight, stateless,
// dependency-free way to send session and event data to HoneyHive.

const {start} = require('@honeyhive/logger');

const sessionId = await start({
    apiKey: MY_HONEYHIVE_API_KEY,
    project: MY_HONEYHIVE_PROJECT_NAME
});

Trace any custom spans using function decorators

The above initialization will auto-capture all interactions with our supported providers.To capture anything else, you can trace any function in your code and see its inputs, outputs, errors, duration, etc. by decorating it as follows.

The following method isn’t compatible with the LangChain callback handlers.

from honeyhive import trace

@trace
def my_function(param1, param2):
    # Code here
    return result

from honeyhive import atrace

@atrace
async def my_function(param1, param2):
    # Code here
    return await async_result

// wrap your function with tracer.traceFunction() {}
// keep the traced function's name the same
const myFunction = tracer.traceFunction()( 
    async function myFunction(query) {
        // some code
        return result;
    }
);

const result = await myFunction("test");

# The honeyhive_logger package provides a lightweight, stateless,
# dependency-free way to send session and event data to HoneyHive.

from honeyhive_logger import log

def my_function(param1, param2):
  # Code here
  return result

result = my_function(param1, param2)

# For maximum insights, learn more about sending data
# in the schema overview documentation at https://docs.honeyhive.ai/schema-overview
log(
  api_key=MY_HONEYHIVE_API_KEY,
  project=MY_HONEYHIVE_PROJECT_NAME,      
  session_id = session_id, # obtained from start(),
  event_name="my_func",
  event_type="tool", 
  inputs = {
    "param1": param1,
    "param2": param2
  },
  outputs = {
    "result": result
  }
)

  // The honeyhive/logger package provides a lightweight, stateless,
  // dependency-free way to send session and event data to HoneyHive.

  const {log} = require('@honeyhive/logger');

  async function myFunction(query: string) {
      // some code
      return result;
  }

  const result = await myFunction(query);


  // For maximum insights, learn more about sending data
  // in the schema overview documentation at https://docs.honeyhive.ai/schema-overview
  const eventId = await log({
      sessionId: sessionId, // obtained from start()
      eventName: "my-func",
      eventType: "tool",
      inputs: {
          query: query
      },
      outputs: {
          result: result
      }
  });    

View the trace

Now that you have successfully traced your session, you can review it in the platform. Navigate to Log Store and click to view any trace.

Learn more

Observability Tutorial

An end-to-end tutorial for tracing a complex RAG application with the tracer.

Tracer Troubleshooting

Learn how to troubleshoot common issues with our tracers.

Enriching Traces

How to add feedback, metrics, metadata, and more to traces.

Data Model Overview

Learn how HoneyHive’s core data model works.

​Logging a Trace

​View the trace

​

​Learn more

Observability Tutorial

Tracer Troubleshooting

Enriching Traces

Data Model Overview

Logging a Trace

View the trace

Learn more