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

# Sediment

> Sediment is an open-source, self-hosted evidence store for coding agents.

Sediment is an open-source, self-hosted evidence store for coding agents.

Capture model calls, code changes, developer decisions, and check results on
your infrastructure.

* [Evaluate agent work](/operate/measure-agent-work): compare models by
  recorded decisions, code retention, and automated checks.
* [Reuse context](/operate/resume-with-evidence): give agents selected
  messages, tool calls, and results from a previous Session.
* [Build training datasets](/exports/training-exports): export examples
  for fine-tuning, preference training, and reinforcement learning.

<Columns cols={2}>
  <Card title="Quickstart" icon="bolt" href="/quickstart">
    Capture your first agent session on one machine in about five minutes.
  </Card>

  <Card title="How it works" icon="book-open" href="/how-sediment-works">
    One evidence store for evaluation, agent context, and training datasets.
  </Card>

  <Card title="Evaluate agent work" icon="chart-line" href="/operate/measure-agent-work">
    Compare model outcomes, code retention, and rework.
  </Card>

  <Card title="Reuse context" icon="rotate" href="/operate/resume-with-evidence">
    Select captured evidence for an agent continuing a task.
  </Card>

  <Card title="Deploy" icon="server" href="/operate/deploy">
    Gateway capture, forge webhooks, per-machine setup, and verification.
  </Card>

  <Card title="Build training datasets" icon="database" href="/exports/training-exports">
    DPO, SFT, diff-SFT, Recovery, and RLVR rows.
  </Card>
</Columns>