> ## Documentation Index > Fetch the complete documentation index at: https://tracevell.com/docs/llms.txt > Use this file to discover all available pages before exploring further. # Agent and LLM Observability Integrations Tracevell is an open-source, OpenTelemetry-native observability platform for AI agents. It auto-instruments LLM provider SDKs, agent frameworks, and browser agents, so you trace, debug, and monitor every call without changing your code. Initialize Tracevell once and calls to OpenAI, Anthropic, LangChain, the Vercel AI SDK, and the rest of the integrations below are captured with full detail: prompts, responses, tool calls, tokens, latency, and cost. Each integration page below has setup steps for TypeScript and Python (where both are supported) and shows what Tracevell captures. Pick the SDK or framework you use. ## LLM provider SDKs Auto-instrumented provider clients. Initialize Tracevell and every model call is traced. > [card] [OpenAI](https://tracevell.com/docs/integrations/openai.md) > > Trace the OpenAI SDK in JavaScript and Python. > [card] [Anthropic](https://tracevell.com/docs/integrations/anthropic.md) > > Trace the Anthropic SDK in JavaScript and Python. > [card] [Gemini](https://tracevell.com/docs/integrations/gemini.md) > > Trace the Google Gemini SDK. > [card] [Cohere](https://tracevell.com/docs/integrations/cohere.md) > > Trace the Cohere SDK. > [card] [LiteLLM](https://tracevell.com/docs/integrations/litellm.md) > > Trace calls routed through LiteLLM. > [card] [OpenRouter](https://tracevell.com/docs/integrations/openrouter.md) > > Trace any model called through OpenRouter. > [card] [TypeSafe AI](https://tracevell.com/docs/integrations/typesafe.md) > > Trace typed Jev decisions: routing, guardrails, and scoring. ## Agent frameworks Multi-step agents, tool calls, and subagents render as a transcript instead of a raw span tree. > [card] [Vercel AI SDK](https://tracevell.com/docs/integrations/vercel-ai-sdk.md) > > Trace `generateText` and `streamText` in Next.js and Node.js. > [card] [eve](https://tracevell.com/docs/integrations/eve.md) > > Trace eve durable agents, their turns, and tool calls. > [card] [LangChain / LangGraph](https://tracevell.com/docs/integrations/langchain.md) > > Trace LangChain chains and LangGraph graphs. > [card] [deepagents](https://tracevell.com/docs/integrations/deepagents.md) > > Trace LangChain deepagents and their subagents. > [card] [Mastra](https://tracevell.com/docs/integrations/mastra.md) > > Trace Mastra agents and workflows in TypeScript. > [card] [Pydantic AI](https://tracevell.com/docs/integrations/pydantic-ai.md) > > Trace Pydantic AI agents and tool calls. > [card] [Microsoft Agent Framework](https://tracevell.com/docs/integrations/microsoft-agent-framework.md) > > Trace Microsoft Agent Framework agents, tools, and workflows. > [card] [OpenAI Agents SDK](https://tracevell.com/docs/integrations/openai-agents-sdk.md) > > Trace OpenAI Agents, handoffs, and tools. > [card] [Google ADK](https://tracevell.com/docs/integrations/google-adk.md) > > Trace Google ADK agents, Gemini calls, and tools. > [card] [OpenHands SDK](https://tracevell.com/docs/integrations/openhands-sdk.md) > > Trace OpenHands software agents. > [card] [Temporal](https://tracevell.com/docs/integrations/temporal.md) > > Trace agent activities across Temporal workflows. ## Coding agents > [card] [Claude Code](https://tracevell.com/docs/integrations/claude-code.md) > > Trace Claude Code sessions with a one-command plugin install. > [card] [Codex](https://tracevell.com/docs/integrations/codex.md) > > Trace OpenAI Codex CLI sessions with a one-command plugin install. > [card] [Pi](https://tracevell.com/docs/integrations/pi.md) > > Trace Pi coding agent runs with a one-command extension install. > [card] [Claude Agent SDK](https://tracevell.com/docs/integrations/claude-agent-sdk.md) > > Trace Claude Agent SDK runs and subagents. > [card] [OpenCode](https://tracevell.com/docs/integrations/opencode.md) > > Trace OpenCode coding agent sessions. ## Browser and computer-use agents Browser session recordings sync with the trace: scrub the video and the trace scrolls with it. > [card] [Browser Use](https://tracevell.com/docs/integrations/browser-use.md) > > Trace Browser Use agents with synced session recordings. > [card] [Stagehand](https://tracevell.com/docs/integrations/stagehand.md) > > Trace Stagehand browser automation. > [card] [Skyvern](https://tracevell.com/docs/integrations/skyvern.md) > > Trace Skyvern browser agents. > [card] [Kernel](https://tracevell.com/docs/integrations/kernel.md) > > Trace computer-using agents running on Kernel. > [card] [Playwright](https://tracevell.com/docs/integrations/playwright.md) > > Trace Playwright browser sessions. > [card] [Puppeteer](https://tracevell.com/docs/integrations/puppeteer.md) > > Trace Puppeteer browser sessions. ## Eval frameworks Run your existing benchmarks and see every trial in Tracevell as a scored datapoint with its full trace. > [card] [Harbor](https://tracevell.com/docs/integrations/harbor.md) > > Send Harbor jobs to Tracevell as evaluations, with verifier rewards as scores and a trace per trial. ## Already on another platform? > [card] [Langfuse](https://tracevell.com/docs/integrations/langfuse.md) > > Keep your Langfuse instrumentation and send the same spans to Tracevell at the same time, so you can compare both without rewriting code. ## Don't see your SDK? Tracevell is OpenTelemetry-native, so it traces any LLM call or function you wrap yourself. Use the `@observe` decorator or create manual spans, and point any OTel-emitting code at Tracevell. See [Tracing Structure](https://tracevell.com/docs/tracing/structure/overview.md) or jump to [observe usage](https://tracevell.com/docs/tracing/structure/observe-decorator.md). ## Once your traces land > [card] [View traces](https://tracevell.com/docs/platform/viewing-traces.md) > > Tracevell extracts inputs, LLM outputs, and tool calls into a transcript view, so you read the conversation instead of a span tree. > [card] [Track outcomes with Signals](https://tracevell.com/docs/signals/introduction.md) > > Define signals in plain language to track the outcomes and failures you care about across every run. > [card] [Query across traces](https://tracevell.com/docs/platform/sql-editor.md) > > Run read-only SQL over your traces in the editor, the SQL API, the [CLI](https://tracevell.com/docs/platform/cli.md), or the [MCP server](https://tracevell.com/docs/platform/mcp.md). > [card] [Tracing structure](https://tracevell.com/docs/tracing/structure/overview.md) > > Use `observe`, sessions, metadata, and tags for deeper control over what gets captured. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.