메뉴
用 Cursor Agent（agent.cmd）进行“分步指挥、逐步验收、失败续跑”的工程改造流程，适合最小侵入式需求落地。
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
Analyze Android accessibility trace logs and convert them into actionable permission-flow insights, click-step reconstruction, and builder-ready automation rules. Use when a user pastes trace content containing AccessibilityEvent logs, windows snapshots, node trees, permission dialogs, ROM-specific permission flows (Xiaomi/HyperOS, HONOR/MagicOS, ZTE/AOSP, etc.), or asks to identify what was clicked, reconstruct authorization steps, summarize a permission flow, or turn a trace into builder/automation logic.
Use the Agent Browser CLI (`agent-browser`) for web automation when the user explicitly asks for "agent-browser/agent-broswer" or wants browser actions executed through that CLI instead of OpenClaw browser tool. Includes install checks, session/profile usage, and stable snapshot-ref workflow.
Configure OpenClaw multi-agent communication with sessions_send routing, allowlists, sender tags, and async ack rules. Use when setting up or fixing agent-to-agent messaging, cross-session relays, or ambiguous sender identity.
Configure OpenClaw multi-agent communication with sessions_send routing, allowlists, sender tags, and async ack rules. Use when setting up or fixing agent-to-agent messaging, cross-session relays, or ambiguous sender identity.
Execute Android accessibility automation from natural language goals (open app, navigate UI, search, input text, and submit) using semantic UI understanding and stepwise planning with safety checks. Use when users ask to control Android apps by intent rather than fixed coordinates, especially for flows like entering search pages first and then typing.
Return a reverse-chronological shortlist of recent conversation records (messages), not sessions. Use when the user asks for 最近N条交流记录/聊天记录, including records across historical sessions, then drill into any selected item.
SOC 직업 분류 기준
| name | cursor-cli-step-orchestrator |
| description | 用 Cursor Agent（agent.cmd）进行“分步指挥、逐步验收、失败续跑”的工程改造流程，适合最小侵入式需求落地。 |
%LOCALAPPDATA%\cursor-agent\agent.cmd
& '%LOCALAPPDATA%\cursor-agent\agent.cmd' \
-p --trust --force \
--workspace '<PROJECT_PATH>' \
"Step N only: <单步目标+边界+验收条件>. Print changed files and concise summary."
说明：
-p：非交互打印模式，便于脚本编排--trust：信任当前 workspace，减少确认阻塞--force：允许执行必要命令（谨慎使用）Step N only：强约束只做当前步，避免越界改动每步结束必须要求 agent 输出：
& '%LOCALAPPDATA%\cursor-agent\agent.cmd' -p --trust --force --workspace '<PROJECT_PATH>' "Step 3 only: implement TrackForegroundService with AMapLocationClient, 60s sampling, START_STICKY, notification channel, cleanup in onDestroy. Keep changes minimal and compile-friendly. Print changed files and summary."