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

# bryel

> OpenTelemetry-native observability for LLM apps and agents — ship traces from any framework to bryel.

Drop in the SDK and every model call, tool call, token count, and cost lands in bryel as OpenInference traces. From there, attach feedback to any generation, turn real runs into datasets, and explore it all — including from a coding agent over the remote MCP.

## Start here

<CardGroup cols={2}>
  <Card title="Quickstart" icon="rocket" href="/quickstart">
    Send your first trace in a few minutes.
  </Card>

  <Card title="SDKs" icon="code" href="/sdk/typescript">
    TypeScript, Python, Browser, or the HTTP API.
  </Card>

  <Card title="Concepts" icon="book-open" href="/concepts">
    Traces, intents, feedback, and datasets.
  </Card>

  <Card title="Coding agents" icon="bot" href="/guides/coding-agents">
    Connect a coding agent over the remote MCP.
  </Card>
</CardGroup>

## What you get

<CardGroup cols={2}>
  <Card title="Any framework via OTel" icon="workflow">
    An OTLP exporter and span processor at the core. Use the Vercel AI SDK adapter, an OpenInference instrumentor, or your own OpenTelemetry spans.
  </Card>

  <Card title="Traces to intents to feedback" icon="layers">
    Capture full agent runs, then attach 👍/👎, scores, and corrections to any generation.
  </Card>

  <Card title="Datasets from production" icon="database">
    Turn real traces and feedback into datasets you can curate and reuse.
  </Card>

  <Card title="Remote MCP for coding agents" icon="bot">
    Give coding agents direct access to your traces through a remote MCP server.
  </Card>
</CardGroup>