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

# Armature: Automated Testing for MCP Servers

> Armature continuously tests your MCP servers with scheduled AI-agent workflows, tool health monitors, and an agent-facing repair API.

Armature helps you catch MCP server regressions before your customers do. Connect your MCP server, define test workflows with natural-language criteria, and Armature runs them automatically on a schedule — evaluating every tool call against your criteria and alerting you the moment something breaks.

<CardGroup cols={2}>
  <Card title="Quick Start" icon="rocket" href="/quickstart">
    Connect your first MCP server and run your first test workflow in minutes.
  </Card>

  <Card title="Core Concepts" icon="book" href="/concepts">
    Understand workflows, runs, tool monitors, and coverage reports.
  </Card>

  <Card title="MCP API" icon="plug" href="/mcp-api/overview">
    Use Armature's agent-facing MCP endpoint to triage, repair, and rerun workflows.
  </Card>

  <Card title="Alerts" icon="bell" href="/alerts/slack">
    Get notified on Slack or email when a workflow fails or a tool monitor fires.
  </Card>
</CardGroup>

## How Armature works

<Steps>
  <Step title="Connect an MCP server">
    Add your MCP server's URL and credentials. Armature probes it live to discover available tools.
  </Step>

  <Step title="Create a workflow">
    Write a natural-language prompt and define observable criteria — the outcomes your agent must produce.
  </Step>

  <Step title="Schedule and run">
    Set a schedule (or trigger manually). Armature dispatches an AI agent against your MCP server and evaluates the result.
  </Step>

  <Step title="Monitor and repair">
    View pass/fail results, inspect tool-call traces, and use the MCP repair API to diagnose and fix failing workflows.
  </Step>
</Steps>

<CardGroup cols={3}>
  <Card title="Workflows" icon="layers" href="/workflows/overview">
    Scheduled AI-agent tests with evaluation criteria.
  </Card>

  <Card title="Tool Monitors" icon="activity" href="/mcp-servers/tool-monitors">
    Lightweight pings that track per-tool uptime.
  </Card>

  <Card title="Coverage Reports" icon="chart-bar" href="/mcp-servers/coverage">
    See which tools are exercised and which need more tests.
  </Card>
</CardGroup>