Menü
Run lifecycle, streaming, follow-ups, and conversation introspection in the Cursor SDK. Use when consuming agent output, handling progress events, or chaining turns.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Basierend auf der SOC-Berufsklassifikation
The `agent` CLI — install, interactive modes (Agent / Plan / Ask), headless `-p` mode for CI, GitHub Actions integration, and how it relates to the TypeScript SDK. Use when scripting Cursor in shell pipelines or workflows instead of (or alongside) the SDK.
Cloud runtime for Cursor SDK — dedicated VMs, repo-bound runs, auto-PR creation, artifacts, and reattachment. Use when running long, review-oriented, or repo-modifying agents off the local machine.
CursorAgentError class hierarchy, isRetryable, and recovery patterns. Use when classifying SDK failures, building retry logic, or surfacing actionable messages.
File-based agent loop hooks via `.cursor/hooks.json` — pre/post commands, governance gates. Use when enforcing policy without modifying agent prompts.
Configure MCP servers on Cursor SDK agents — stdio subprocess, HTTP, and SSE transports, inline vs file-based config, and per-run overrides. Use when extending an agent's tools beyond the built-ins.
Model discovery and selection on Cursor SDK — Cursor.models.list, ModelSelection, parameters (e.g. thinking levels), per-run sticky overrides.
| name | cursor-runs |
| description | Run lifecycle, streaming, follow-ups, and conversation introspection in the Cursor SDK. Use when consuming agent output, handling progress events, or chaining turns. |
| metadata | {"internal":false,"source":"https://cursor.com/docs/sdk/typescript"} |
Everything you do after Agent.create flows through a Run. This skill covers how to issue work, stream events, await results, cancel, and inspect conversation history.
SDKMessage eventsagent.send(message, options) and SendOptionsforceid, status, result, git, model, durationMsawait run.wait() and RunResultrun.cancel() and cancellation semanticsonDidChangeStatus listenerrun.supports(op) and unsupportedReasonfor await over run.stream()onDelta updates: text, thinking, tool calls, tokens, steps, summaries, shellonStep step-level callbackrun.conversation() returns ConversationTurn[]agentConversationTurn vs shellConversationTurn, step types