export const type = "cursor"; export const label = "Cursor CLI (local)"; export const DEFAULT_CURSOR_LOCAL_MODEL = "gpt-5"; export const models = [ { id: DEFAULT_CURSOR_LOCAL_MODEL, label: DEFAULT_CURSOR_LOCAL_MODEL }, { id: "gpt-5-mini", label: "gpt-5-mini" }, { id: "sonnet-4", label: "sonnet-4" }, { id: "sonnet-4-thinking", label: "sonnet-4-thinking" }, ]; export const agentConfigurationDoc = `# cursor agent configuration Adapter: cursor Use when: - You want Paperclip to run Cursor Agent CLI locally as the agent runtime - You want Cursor chat session resume across heartbeats via --resume - You want structured stream output in run logs via --output-format stream-json Don't use when: - You need webhook-style external invocation (use openclaw or http) - You only need one-shot shell commands (use process) - Cursor Agent CLI is not installed on the machine Core fields: - cwd (string, optional): default absolute working directory fallback for the agent process (created if missing when possible) - instructionsFilePath (string, optional): absolute path to a markdown instructions file prepended to the run prompt - promptTemplate (string, optional): run prompt template - model (string, optional): Cursor model id (for example gpt-5) - mode (string, optional): Cursor execution mode passed as --mode (plan|ask) - command (string, optional): defaults to "agent" - extraArgs (string[], optional): additional CLI args - env (object, optional): KEY=VALUE environment variables Operational fields: - timeoutSec (number, optional): run timeout in seconds - graceSec (number, optional): SIGTERM grace period in seconds Notes: - Runs are executed with: agent -p --output-format stream-json ... - Prompts are passed as a final positional argument. - Sessions are resumed with --resume when stored session cwd matches current cwd. `;