video-multimodal-generate
Skill33 starsChanged 5 months agoMITView on GitHub
使用 meitu-cli 做多模态混合驱动生成视频(音频/视频/图片/文字任意组合)。当用户说配合音乐生成、参考视频运镜/风格/节奏、音频驱动视频、多素材混合生成时触发。
- Reads credentials
In this repository55 other files it gives its agents.
- meitu-toolsSkill
- SKILL.mdSkill
- audio-music-generateSkill
- audio-song-generateSkill
- image-background-replaceSkill
- image-cutoutSkill
- image-denoise-enhanceSkill
- image-editSkill
- image-element-removeSkill
- image-face-swapSkill
- image-grid-splitSkill
- image-id-photo-generateSkill
- image-lowlight-enhanceSkill
- image-outfit-swapSkill
- image-portrait-generateSkill
- image-poster-generateSkill
- image-searchSkill
- image-style-transferSkill
- image-superres-enhanceSkill
- image-text-replaceSkill
- image-to-videoSkill
- image-transformSkill
- meitu-beautySkill
- meitu-carouselSkill
- meitu-cutoutSkill
- meitu-game-2d-assetsSkill
- meitu-id-photoSkill
- meitu-image-adaptSkill
- meitu-image-fixSkill
- meitu-posterSkill
- meitu-product-swapSkill
- meitu-product-viewSkill
- meitu-stickersSkill
- meitu-upscaleSkill
- meitu-video-danceSkill
- meitu-visual-meSkill
- text-code-editSkill
- text-codeSkill
- text-generateSkill
- text-to-imageSkill
- text-to-videoSkill
- video-audio-addSkill
- video-canvas-expandSkill
- video-content-replaceSkill
- video-denoise-enhanceSkill
- video-effect-applySkill
- video-element-removeSkill
- video-framerate-enhanceSkill
- video-logo-addSkill
- video-lowlight-enhanceSkill
- video-motion-transferSkill
- video-quality-enhanceSkill
- video-resolution-upscaleSkill
- video-stitchSkill
- video-to-gifSkill
The file
Files
Video Multimodal Generate
Overview
调用 meitu video-multimodal-generate 基于音频/视频/图片/文字任意组合做多模态混合驱动视频生成。适合音乐配视频、参考视频运镜/节奏、多素材混合生成。
不处理:纯文字生成视频、仅图片生成视频(无视频无音频)、动作迁移。
API Mapping
| DAG 工具 | 后端 API |
| video-multimodal-generate | video_multimodal_generate |
Dependencies
- meitu-cli:
>=2.0.6 - 凭证:CONFIG AKSK →
meitu tools update;EXEC AKSK → 跑命令 - 环境变量:
MEITU_OPENAPI_TOOL_TASK_MODE=command
路径别名:
$VISUAL={OPENCLAW_HOME}/workspace/visual/
Core Workflow
Preflight → Execute → DeliverPreflight
meitu --version≥ 2.0.6;meitu auth verify --json- 确认已跑过 Config Phase;
MEITU_OPENAPI_TOOL_TASK_MODE=command - output_dir 解析:openclaw.yaml →
./output/| else →$VISUAL/output/video-multimodal-generate/;mkdir -p {output_dir}
Execute
触发信号/路由
触发条件(任一成立):
- 用户上传音频 → 必须走本工具(独有能力)
- 仅视频无图片 → 运镜/风格/节奏参考
- 用户想参考视频的运镜/风格/节奏(不是模仿人体动作)
- 复杂多素材混合(图片+视频+音频)
决策顺序:
- 至少一类多媒体素材;
prompt缺失 → 追问 - 素材总数 ≤12(图片+视频+音频合计)
- 唯一路由 →
video_multimodal_generate
有图+视频但意图不明(运镜 vs 动作模仿)→ 默认走本工具。
参数定义
| 参数 | 类型 | 必填 | 范围 | 默认值 | 说明 |
prompt | STRING | 是 | -- | -- | 文字描述;缺失 → "请描述你想生成的视频内容" |
video_duration | NUMBER | 是 | 4–15 / -1(智能) | -1 | 视频时长(秒) |
ratio | STRING | 是 | adaptive/16:9/4:3/1:1/3:4/9:16/21:9 | adaptive | 画面比例 |
image_list | ARRAY | 否 | ≤9 张 | -- | 图片 URL 列表 |
reference_video_list | ARRAY | 否 | ≤3 个,总时长 ≤15s | -- | 参考视频 URL 列表 |
reference_audio_list | ARRAY | 否 | ≤3 个,总时长 ≤15s | -- | 音频 URL 列表 |
sound | BOOL | 否 | true / false | true | 是否生成音频 |
resolution | STRING | 否 | 480p / 720p | 720p | 视频分辨率 |
至少提供一类多媒体素材;图片+视频+音频合计 ≤12。
工具调用
meitu video-multimodal-generate \
--skill_name skill_video-multimodal-generate \
--prompt "{prompt}" \
--video_duration -1 \
--ratio adaptive \
--image_list '["{url1}"]' \
--reference_video_list '["{video_url}"]' \
--reference_audio_list '["{audio_url}"]' \
--sound true --resolution 720p \
--json --download-dir {output_dir}错误降级
| 场景 | 处理 |
prompt 缺失但有素材 | 提示"请描述你想生成的视频内容" |
| 素材总数 >12 | 提示输入限制,建议精简素材 |
video_duration 与音频时长冲突 | 优先匹配音频时长,告知用户 |
| 部分素材不可访问 | 静默过滤,按剩余素材继续执行 |
| 调用失败 | 降级模型版本重试 1 次 → 仍失败返回错误 |
| 超时 | 降低 duration / size 重试 1 次 → 仍失败返回错误 |
| 内容合规拦截 | 直接返回合规提示,不重试 |
| 无任何多媒体素材 | 不属于本工具范畴 |
Deliver
解析 --json → downloaded_files[0].saved_path 或 media_urls[0]。
mv {file} {output_dir}/{YYYY-MM-DD}_{name}_multimodal.mp4
Output
- 格式:MP4(默认 720p,可选 480p)
- 命名:
{YYYY-MM-DD}_{descriptive-name}_multimodal.mp4 - 位置:项目 →
./output/;一次性 →$VISUAL/output/video-multimodal-generate/
基线 Task ID
t_mt1a3i5n7bd3dd7a38-c263-48f7-ac18-bb8573c364b2Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

