> ## Documentation Index
> Fetch the complete documentation index at: https://docs.honeyhive.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Overview

> Connect HoneyHive to the agent frameworks, model providers, and tools in your AI stack

HoneyHive captures traces, metrics, and evaluations from the tools you already use to build AI applications. The integrations below give you framework-specific setup, but they aren't a hard boundary: HoneyHive's automatic tracing builds on [OpenTelemetry semantic conventions](/v2/tracing/introduction), so any OTel-instrumented library can send data to HoneyHive even without a dedicated guide.

New to tracing? Start with the [Quickstart](/v2/introduction/tracing-quickstart).

***

## Agent Frameworks & Platforms

Frameworks and visual builders for composing agents, chains, and multi-step workflows.

<CardGroup cols={3}>
  <Card title="Google ADK" icon="robot" href="/v2/integrations/google-adk">
    Google's Agent Development Kit
  </Card>

  <Card title="AWS Strands" icon="robot" href="/v2/integrations/strands">
    AWS Strands Agents
  </Card>

  <Card title="Semantic Kernel" icon="robot" href="/v2/integrations/semantic-kernel">
    Microsoft's agent orchestration SDK
  </Card>

  <Card title="AutoGen" icon="robot" href="/v2/integrations/autogen">
    Microsoft's multi-agent framework
  </Card>

  <Card title="LangGraph" icon="robot" href="/v2/integrations/langgraph">
    Stateful agent graphs from LangChain
  </Card>

  <Card title="CrewAI" icon="robot" href="/v2/integrations/crewai">
    Role-based multi-agent crews
  </Card>

  <Card title="OpenAI Agents" icon="robot" href="/v2/integrations/openai-agents">
    OpenAI Agents SDK
  </Card>

  <Card title="LangChain" icon="robot" href="/v2/integrations/langchain">
    Chains and LLM application building blocks
  </Card>

  <Card title="LlamaIndex" icon="robot" href="/v2/integrations/llamaindex">
    Data framework for RAG and agents
  </Card>

  <Card title="Pydantic AI" icon="robot" href="/v2/integrations/pydantic-ai">
    Type-safe agents built on Pydantic
  </Card>

  <Card title="Claude Agent SDK" icon="robot" href="/v2/integrations/claude-agent-sdk">
    Anthropic's agent-building SDK
  </Card>

  <Card title="Cursor SDK" icon="robot" href="/v2/integrations/cursor-sdk">
    Cursor's programmatic agent SDK
  </Card>

  <Card title="DSPy" icon="robot" href="/v2/integrations/dspy">
    Declarative language model programs
  </Card>

  <Card title="n8n" icon="robot" href="/v2/integrations/n8n">
    Visual AI workflow and agent builder
  </Card>

  <Card title="Copilot Studio" icon="robot" href="/v2/integrations/copilot-studio">
    Microsoft's low-code agent platform
  </Card>
</CardGroup>

***

## Model Providers

Capture LLM requests and responses directly from provider SDKs.

<CardGroup cols={3}>
  <Card title="OpenAI" icon="sparkles" href="/v2/integrations/openai">
    GPT models via the OpenAI API
  </Card>

  <Card title="Anthropic" icon="sparkles" href="/v2/integrations/anthropic">
    Claude models via the Anthropic API
  </Card>

  <Card title="Google Gemini" icon="sparkles" href="/v2/integrations/gemini">
    Gemini models via the Google API
  </Card>

  <Card title="Azure OpenAI" icon="sparkles" href="/v2/integrations/azure_openai">
    OpenAI models on Azure
  </Card>

  <Card title="AWS Bedrock" icon="sparkles" href="/v2/integrations/aws_bedrock">
    Foundation models on AWS Bedrock
  </Card>
</CardGroup>

***

## AI Gateways

Route requests through a gateway and forward traces to HoneyHive.

<CardGroup cols={3}>
  <Card title="LiteLLM" icon="route" href="/v2/integrations/litellm">
    Unified proxy across 100+ models
  </Card>

  <Card title="Portkey" icon="route" href="/v2/integrations/portkey">
    AI gateway and observability layer
  </Card>

  <Card title="TrueFoundry" icon="route" href="/v2/integrations/truefoundry">
    LLM gateway and deployment platform
  </Card>
</CardGroup>

***

## Coding Agents

Trace autonomous coding agents and their tool use.

<CardGroup cols={3}>
  <Card title="Claude Code" icon="terminal" href="/v2/integrations/claude-code">
    Anthropic's agentic coding CLI
  </Card>

  <Card title="Devin" icon="terminal" href="/v2/integrations/devin">
    Cognition's autonomous software engineer
  </Card>
</CardGroup>

***

## Evaluation

Run third-party evaluation frameworks against HoneyHive-traced agents.

<CardGroup cols={3}>
  <Card title="ASSERT" icon="shield-check" href="/v2/integrations/assert">
    Microsoft's spec-driven behavior evals
  </Card>
</CardGroup>

***

## Don't see your tool?

HoneyHive's tracing is built on OpenTelemetry, so most libraries work without a dedicated integration. Follow the [Tracing Introduction](/v2/tracing/introduction) to instrument any framework with the standard SDK setup, or [reach out](mailto:support@honeyhive.ai) if you'd like us to prioritize a guide.