sync: update ai-chat, veo-video, seedance-video from Docs (8430b01) [auto-sync]

[Copilot]

Syncs three SKILL.md files against the latest AceDataCloud/Docs OpenAPI specs. Three skills had drifted: model lists out of date, and Seedance's entire request schema changed.

ai-chat

• OpenAI: add gpt-5.4, gpt-5.4-pro, gpt-5.2, gpt-5.1, gpt-5-nano, o3, o3-mini, o3-pro, o4-mini
• Claude: add claude-opus-4-1-20250805, claude-opus-4-20250514, claude-3-7-sonnet-20250219, claude-3-5-haiku-20241022, claude-3-5-sonnet-20240620, claude-3-haiku-20240307, claude-3-sonnet-20240229
• Gemini: replace gemini-1.5-pro/flash → gemini-2.0-flash, gemini-2.5-flash/pro, gemini-3-flash-preview, gemini-3.0-pro, gemini-3.1-pro
• DeepSeek: add deepseek-v3.2-exp
• Grok: add grok-4-1-fast, grok-4-1-fast-non-reasoning, grok-2-vision; remove grok-3-fast

veo-video

Model identifiers changed (no more hyphens), plus new fast variants and a required action field:

// Before
{ "model": "veo-3", "image_url": "https://..." }

// After
{ "action": "text2video", "model": "veo3" }
{ "action": "image2video", "model": "veo2", "image_urls": ["https://..."] }
{ "action": "get1080p", "video_id": "..." }

New models: veo2-fast, veo3-fast, veo31-fast, veo31-fast-ingredient.

seedance-video

Major request schema change — flat prompt/image_url replaced by a typed content array; model IDs renamed to doubao-seedance-*:

// Before
{ "model": "seedance-1.5-pro", "prompt": "...", "image_url": "...", "service_tier": "premium" }

// After
{
  "model": "doubao-seedance-1-5-pro-251215",
  "content": [
    { "type": "image_url", "role": "first_frame", "image_url": { "url": "..." } },
    { "type": "text", "text": "..." }
  ],
  "service_tier": "default"
}

New parameters: ratio, frames, camerafixed, watermark, generate_audio (1.5-pro only), return_last_frame. service_tier values changed from standard/premium → default/flex.