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

# Traces

> Inspect every run, span, token and cost.

The **Traces** page lists recent traces for the project and lets you drill into
any one.

## Trace list

Each row summarizes a trace: name (root span), start time, duration, span count,
error count, services, total tokens, and total cost. Filter by `user_id`,
`session_id`, or search by `trace_id`.

## Trace detail

Opening a trace shows:

* **Span waterfall** — the spans laid out by time, so you can see planning,
  retrieval, tool calls and generations in sequence.
* **Per-span detail** — model, token breakdown (input / output / cache), cost,
  latency, status, and the input/output payloads.
* **Scores** — any eval or annotation scores attached to the trace.

## Where traces come from

Anything that exports to the [OTLP endpoint](/integrations/otlp):

* [Claude Agent SDK](/integrations/claude-agent-sdk) via `trace()`
* [Vercel AI SDK](/integrations/vercel-ai-sdk) via OpenTelemetry
* [Claude Code](/integrations/claude-code) sessions via `/v1/logs`

<Tip>
  Turn a real trace into a regression case by adding it to a
  [dataset](/app/datasets).
</Tip>