Global Rank · of 601 Skills
byted-las-video-inpaint AI Agent Skill
View Source: bytedance/agentkit-samples
SafeInstallation
npx skills add bytedance/agentkit-samples --skill byted-las-video-inpaint 14
Installs
LAS 视频修复(las_video_inpaint)
本 Skill 基于以下两个接口,封装 submit/poll 异步调用流程。
POST https://operator.las.cn-beijing.volces.com/api/v1/submit提交修复任务POST https://operator.las.cn-beijing.volces.com/api/v1/poll轮询任务状态并拿到修复后视频
你需要准备什么
LAS_API_KEY:优先从环境变量读取;也支持放在当前目录的env.sh- Operator Region(二选一):
- 环境变量:
LAS_REGION(推荐)/REGION/region,取值cn-beijing(默认)或cn-shanghai - 或命令中通过
--region cn-shanghai指定
- 环境变量:
video_url:可下载的视频地址(http/https或tos://bucket/key)output_tos_path:修复结果写入的 TOS 路径前缀(建议传目录前缀)。实际输出以inpainted_video_path为准
参数与返回字段(详细版)
完整速查见:
推荐使用方式
本 Skill 自带可执行脚本:scripts/skill.py。
下面示例默认你位于该 Skill 目录(与 SKILL.md 同级),因此命令使用相对路径 scripts/skill.py。
1) 仅提交(返回 task_id)
python3 scripts/skill.py submit \
--video-url "tos://bucket/input.mp4" \
--output-tos-path "tos://bucket/output/" \
--targets watermark \
--targets subtitle \
--region cn-beijing \
--out submit.json2) 查询任务状态(poll)
python3 scripts/skill.py poll task-xxx \
--region cn-beijing \
--out result.json建议在对话中继续处理其他问题;每隔一段时间(例如 5-10 秒)再 poll 一次,直到 task_status=COMPLETED 后把 data.inpainted_video_path 返回给用户。
关键参数说明
targets:擦除目标类型,常用watermark/subtitle/scrolling_subtitle,默认服务端通常为["watermark","subtitle"]fixed_bboxes:如果你只想修复某些固定区域,可用该参数。坐标以1000x1000为基准,会按实际分辨率缩放return_subtitle_bbox:是否返回字幕 bbox(字符串x1,y1,x2,y2,原视频像素坐标)- 如果你传入的是带
.mp4的路径(看起来像文件),服务端可能仍会在该路径下追加任务目录与文件名;建议使用目录前缀,避免出现...mp4/<task_dir>/<file>.mp4的层级
使用限制(摘要)
- 视频源需可访问(公网/火山内网/TOS)
- 视频时长 ≤ 4 小时,大小 ≤ 30GB
补充:本地文件作为输入输出
las_video_inpaint 输入支持TOS或可下载URL,输出必须写入 TOS 上(最终以 inpainted_video_path 为准)。当用户给的是本地文件路径或希望本地落盘时,按下面规则处理,并配合 byted-tosfile-access 做上传/下载中转。
规则
- 用户输入是本地路径:先用 byted-tosfile-access 上传到 TOS,得到
tos://...再作为--video-url调用本技能。 - 用户输出要求是本地路径:本技能先输出到 TOS(中转),待
COMPLETED后用 byted-tosfile-access 把inpainted_video_path下载到本地。 - 用户输入与输出都要求本地:必须追问用户提供一个“可写的 TOS 中转目录前缀”(例如
tos://bucket/tmp/video_inpaint/),用它同时承载上传的输入与算子输出。 - 用户输入或输出有一项已经是
tos://...且仍需要中转:优先复用该 bucket,并通过“路径改写”生成中转前缀,例如把tos://bucket/a/b/input.mp4改写成tos://bucket/a/b/video_inpaint/(或tos://bucket/a/b/tmp/video_inpaint/),避免覆盖原文件。
Installs
Security Audit
View Source
bytedance/agentkit-samples
More from this source
Power your AI Agents with
the best open-source models.
Drop-in OpenAI-compatible API. No data leaves Europe.
Explore Inference APIGLM
GLM 5
$1.00 / $3.20
per M tokens
Kimi
Kimi K2.5
$0.60 / $2.80
per M tokens
MiniMax
MiniMax M2.5
$0.30 / $1.20
per M tokens
Qwen
Qwen3.5 122B
$0.40 / $3.00
per M tokens
How to use this skill
Install byted-las-video-inpaint by running npx skills add bytedance/agentkit-samples --skill byted-las-video-inpaint in your project directory. Run the install command above in your project directory. The skill file will be downloaded from GitHub and placed in your project.
No configuration needed. Your AI agent (Claude Code, Cursor, Windsurf, etc.) automatically detects installed skills and uses them as context when generating code.
The skill enhances your agent's understanding of byted-las-video-inpaint, helping it follow established patterns, avoid common mistakes, and produce production-ready output.
What you get
Skills are plain-text instruction files — not executable code. They encode expert knowledge about frameworks, languages, or tools that your AI agent reads to improve its output. This means zero runtime overhead, no dependency conflicts, and full transparency: you can read and review every instruction before installing.
Compatibility
This skill works with any AI coding agent that supports the skills.sh format, including Claude Code (Anthropic), Cursor, Windsurf, Cline, Aider, and other tools that read project-level context files. Skills are framework-agnostic at the transport level — the content inside determines which language or framework it applies to.
Chat with 100+ AI Models in one App.
Use Claude, ChatGPT, Gemini alongside with EU-Hosted Models like Deepseek, GLM-5, Kimi K2.5 and many more.