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

# Cursor

> Launch and monitor Cursor Cloud Agents — agent lifecycle, background runs, and artifacts

# Cursor

Shanone's Cursor integration gives your agent 17 tools for controlling Cursor's own Cloud Agents API: launching agents, monitoring runs, sending follow-ups, and retrieving artifacts.

## Getting Started

<Steps>
  <Step title="Generate a Cursor API Key">
    In Cursor, go to [cursor.com/settings](https://www.cursor.com/settings) and create an API Key.
  </Step>

  <Step title="Add it to Shanone">
    Open **Integrations** in the Shanone dashboard, select Cursor, and paste in the API Key.
  </Step>

  <Step title="Retry your request">
    Once saved, `shanone_execute_tool` calls for `cursor_*` tools will succeed.
  </Step>
</Steps>

## Available Tools

Shanone provides **17 tools** for Cursor, organized into these categories:

<AccordionGroup>
  <Accordion title="Cloud Agents">
    `cursor_list_agents`, `cursor_get_agent_status`, `cursor_get_agent_conversation`, `cursor_launch_agent`, `cursor_add_followup`, `cursor_delete_agent`, `cursor_archive_agent`, `cursor_unarchive_agent`
  </Accordion>

  <Accordion title="Runs">
    `cursor_list_runs`, `cursor_get_run`, `cursor_create_run`, `cursor_cancel_run`
  </Accordion>

  <Accordion title="Artifacts">
    `cursor_list_artifacts`, `cursor_download_artifact`
  </Accordion>

  <Accordion title="Account Info">
    `cursor_get_api_key_info`, `cursor_list_models`, `cursor_list_repositories`
  </Accordion>
</AccordionGroup>

## Common Use Cases

<CardGroup cols={2}>
  <Card title="Delegate background coding" icon="bot">
    Launch a Cloud Agent on a repository with a task prompt and poll it to completion
  </Card>

  <Card title="Multi-agent orchestration" icon="git-fork">
    Fan out several agents across repos/branches and collect their conversations for review
  </Card>

  <Card title="Follow-up steering" icon="message-square-plus">
    Send a follow-up instruction to a running agent instead of starting a new one from scratch
  </Card>

  <Card title="Artifact retrieval" icon="download">
    Pull generated diffs, logs, or build outputs from a completed run
  </Card>
</CardGroup>

## Troubleshooting

<AccordionGroup>
  <Accordion title="cursor_launch_agent fails with a repository access error">
    The target repository must be listed in `cursor_list_repositories` for the connected account — Cursor Cloud Agents can only act on repositories the API Key's account actually has access to.
  </Accordion>

  <Accordion title="cursor_get_agent_status stays in a running state indefinitely">
    Long-running agents are expected to take minutes; poll `cursor_get_agent_status` periodically rather than expecting immediate completion, and check `cursor_get_agent_conversation` for intermediate progress.
  </Accordion>

  <Accordion title="cursor_add_followup has no effect">
    Follow-ups can only be sent to agents that are still active (not archived, deleted, or already finished) — check `cursor_get_agent_status` first, and use `cursor_launch_agent` to start fresh if the agent has already completed.
  </Accordion>
</AccordionGroup>