菜单
Identify a real-world object the user is looking at or asking about, using the latest passthrough vision frame and gaze. Use for "what is this?", naming objects on a desk, or labeling things in the room.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
Break a user goal into a small DAG of independent sub-tasks and decide which specialist should own each. Use when a request bundles multiple intents (e.g. "what's this AND the weather AND start a timer").
Look up current information on the web (weather, facts, news, prices) and summarize it. Use for "search for…", "what's the weather", "look up…", or any knowledge question. May delegate per-source summarization to sub-agents.
Show live, rolling captions of speech Jarvis hears — for accessibility, noisy rooms, or meetings — with optional translation. Use for "caption this", "turn on subtitles", "caption what they're saying", or "subtitle this meeting". Triggers: caption, captions, subtitles, transcribe live, what are they saying, accessibility, hard of hearing, meeting transcript.
Draft a message, email, or reply in the user's voice and show it for review before anything is sent. Use for "text Sarah that I'm running late", "draft an email to the team", "reply saying yes", or composing a note to someone. Triggers: text, message, email, reply, draft, tell them, send a note, write to.
Translate a live two-way conversation between two languages in real time, showing both sides on a translator panel. Use for "translate this conversation", "help me talk to them in Spanish", or interpreting back-and-forth speech. Triggers: translate conversation, interpret, talk to them in, live translation, two-way translate, help me communicate.
Play, pause, seek, and adjust audio or video, and react to transport controls. Use for "play some music", "pause", "skip ahead 30 seconds", "turn it up", or "play that video". Triggers: play, pause, resume, stop, skip, seek, rewind, volume, louder, quieter, next track, music, video.
基于 SOC 职业分类
| name | identify-object |
| description | Identify a real-world object the user is looking at or asking about, using the latest passthrough vision frame and gaze. Use for "what is this?", naming objects on a desk, or labeling things in the room. |
| license | MIT |
| metadata | {"agent":"perception-agent","category":"perception","version":"1.0"} |
| allowed-tools | identify_object look read_text |
vision_annotation on the object and narrate via agent.observation.