Update to Visual Sandbox

{
  "description": "Get a new person started with Visual Sandbox over the MCP tools, in ChatGPT, Claude or any other chat client, one step per message. Trigger right after the plugin is installed, when the person says \"set up Visual Sandbox\", \"what can you make\", \"get started\", \"how does this work\", or when the first tool call says the account is not connected. Checks the connection, turns the person's photos, product, logo or website into saved anchors, makes one preview keyframe, a first 5-second video and three voice takes, then opens the canvas. Every step is optional. Uses the MCP tools only, never a shell.\n",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 280
    }
  ],
  "name": "vsb-setup",
  "skill_md_contents": "---\nname: vsb-setup\ndescription: >\n  Get a new person started with Visual Sandbox over the MCP tools, in ChatGPT,\n  Claude or any other chat client, one step per message. Trigger right after\n  the plugin is installed, when the person says \"set up Visual Sandbox\",\n  \"what can you make\", \"get started\", \"how does this work\", or when the first\n  tool call says the account is not connected. Checks the connection, turns\n  the person's photos, product, logo or website into saved anchors, makes one\n  preview keyframe, a first 5-second video and three voice takes, then opens\n  the canvas. Every step is optional. Uses the MCP tools only, never a shell.\nlicense: Proprietary. See LICENSE for the terms.\n---\n\n# Get started with Visual Sandbox\n\nVisual Sandbox runs image, video, audio, 3D and vector models on the person's\naccount. Each run is charged to that account. This skill takes a new person\nfrom \"installed\" to a first video they like.\n\n**One step for each message.** Do one step, show the result, and ask one\nquestion. Every step is optional: when the person wants to skip one, skip it.\nUse the MCP tools only. There is no shell and no CLI on this surface.\n\n## 1. Check the connection\n\nCall `get_balance`.\n\n- **It answers.** The account is connected. Go on.\n- **It refuses with \"Authorization is required\".** Tell the person to open\n  the plugin settings, press Connect and sign in to Visual Sandbox. Stop until\n  they come back.\n- **It refuses with `email_unverified`.** Tell the person to confirm the\n  address from the Visual Sandbox email. Generation waits until they do.\n\nWhen `balance_usd` is zero, say that runs are charged to the account and that\ncredit is added on visualsandbox.com. Do not quote a price for credit, and do\nnot offer to buy it.\n\n## 2. Ask what to make first\n\nSay in one sentence what the plugin makes: videos, ads, voiceovers, product\nphotos, music, logos and pages. Then ask what they want to make first, and\nask for what they have:\n\n- a photo of themselves or of a person in the work,\n- a product photo,\n- a logo or a website,\n- a script or an idea.\n\n## 3. Turn their files into anchors\n\nAn anchor is saved once and used in every later run. Make only the ones that\nthe work needs.\n\n- **A file in the chat** goes to `upload_media` as `file`. In ChatGPT, when\n  nothing is attached, `import_media` opens a card where they pick a file.\n- **A person or a mascot:** a character sheet, then `create_character`. See\n  [`vsb-characters`](../vsb-characters/SKILL.md).\n- **A website or a logo:** a brand theme with `extract_theme`, checked by the\n  person before it is saved. See\n  [`vsb-library-themes`](../vsb-library-themes/SKILL.md).\n- **A product photo:** name it with `update_asset`, such as `acme-mug`, so\n  later runs find it.\n\n## 4. One preview keyframe\n\nMake one image at the shape of the video they want (9:16 for Reels and\nTikTok, 16:9 for YouTube). It is the first frame of the video. Read\n[`vsb-image-prompting`](../vsb-image-prompting/SKILL.md) before the prompt.\nUse the anchors as references.\n\nShow it and ask: \"Is this the look?\" Change it until they agree. Make no\nvideo before they agree.\n\n## 5. A first 5-second video\n\n1. Pick the cheapest video model that takes a first frame. Call\n   `estimate_cost` and say the price.\n2. Run it with the approved keyframe as the first frame, 5 seconds, at the\n   keyframe's shape. Read [`vsb-video`](../vsb-video/SKILL.md) for the model.\n3. In Review mode a run of 10 cents or more waits for the person's press of\n   Generate: show the card, or give the review link. Never press for them.\n\n## 6. Three voices, while the video renders\n\nAsk whether the work needs a voice. When it does:\n\n1. Run `audio_elevenlabs_voice_design` with a description of the voice,\n   `auditions: 3`, and a sample line of 100 to 400 characters in the\n   person's own words. That costs a few cents.\n2. In ChatGPT, call `voice_studio` with the job id. The person hears the\n   three takes and saves one to a character. Elsewhere, give the share link\n   and ask which take they like.\n\n## 7. Offer what is still missing\n\nOffer one of these, the one that fits the work:\n\n- their own likeness, from a photo of them,\n- their own voice, cloned with their consent\n  ([`vsb-clone-voice`](../vsb-clone-voice/SKILL.md)),\n- a brand theme, when the work is for a brand.\n\n## 8. Open the canvas\n\nGive the `sandbox_url`. Every result is on that canvas. Then suggest two or\nthree next prompts that build on what they made, such as \"Make this into a\n10-second ad with a voiceover\" or \"Cut three vertical clips from my podcast\".\n\n## Rules\n\n- Every run costs money. Never run twice to check a result you have:\n  `get_job` returns the image, or four frames of a video.\n- One request is one paid attempt. Ask before another take.\n- Send the same `sandbox_name` on every run in the setup.\n- When a run is refused for want of credit, say the balance is too low, and\n  follow the server instructions about credit.\n"
}