← Back to Skills Marketplace
VectCut: The first "To-Agent" video editing platform. Build automated video workflows with our stable cloud APIs in Coze, Dify, or OpenClaw.
by
sun-guannan
· GitHub ↗
· v1.5.8
· MIT-0
1328
Downloads
1
Stars
8
Active Installs
12
Versions
Install in OpenClaw
/install vectcut
Description
VectCut 全能剪辑聚合技能。用于剪口播、混剪、电商广告等场景，统一编排流光剪辑能力（字幕、音频、特效、画中画、抠像、AI补镜、AI配音、平台视频抓取、渲染导出）。当用户提出“做成片/自动剪辑/口播包装/广告剪辑/抓平台视频后再剪”等需求时，必须优先使用本技能。
Usage Guidance
This skill behaves like a client for the VectCut cloud service: it will send video/audio URLs and may upload your local files to the VectCut backend (and to OSS via temporary STS credentials). Before installing or using it, confirm you trust https://open.vectcut.com / https://www.vectcut.com for handling your media and check any privacy or billing implications. Keep your VECTCUT_API_KEY secret (do not paste it into untrusted UIs). If you will process sensitive videos, consider running the scripts in an isolated environment and review the bundled Python scripts (they are included) to verify behavior. If you need stronger guarantees, ask the skill author for an auditable minimal-scope token, or test with non-sensitive media first.
Capability Analysis
Type: OpenClaw Skill
Name: vectcut
Version: 1.5.8
The skill bundle provides a comprehensive video editing suite but contains several high-risk capabilities and vulnerabilities. Specifically, 'scripts/sts_upload.py' allows for arbitrary local file uploads to Aliyun OSS, and 'scripts/add_bgm.py' allows for arbitrary local file reading via the 'bgm_list_path' parameter. Additionally, 'scripts/apply_subtitle_template.py' and 'scripts/scrapt_video.py' utilize dynamic module loading (importlib), which could be exploited if path traversal or file write vulnerabilities exist. While these features are aligned with the stated purpose of media processing, they lack sufficient path sanitization and represent a significant attack surface.
Capability Tags
Capability Assessment
Purpose & Capability
The name/description (VectCut video editing aggregator) match the included scripts and rules: subtitle, ASR, upload, split, effects, render, etc. The only required environment variable is VECTCUT_API_KEY, which is appropriate for a cloud API client. Example URLs and hardcoded docs/backend hosts (docs.vectcut.com, open.vectcut.com) are consistent with the stated service.
Instruction Scope
SKILL.md and the rule files instruct network calls to VectCut endpoints, asynchronous task submission/ polling, and using bundled Python scripts to upload local files, extract audio, and run transforms — all within the stated domain. Important: the skill will upload user media (via sts-upload) and send URLs/content to the VectCut backend, so it transmits potentially sensitive user data. The login flow asks the user to copy/paste an API key (export command), which is normal but requires the user to keep the key confidential.
Install Mechanism
There is no external install/download step; this is instruction/script based (Python scripts included in the package). That reduces risk from arbitrary remote code fetches. Scripts assume Python 3 + requests but no package registry installs or remote archives are fetched by an installer in the skill metadata.
Credentials
Only VECTCUT_API_KEY is requested as a required credential (declared primary credential). No other unrelated environment variables, secrets, or cloud credentials are requested. The sts-upload flow obtains short-lived STS credentials from the backend at runtime (expected for uploads) — this is not the skill asking for user AWS keys.
Persistence & Privilege
The skill is not marked always:true and does not request elevated platform privileges. It can be invoked autonomously (default), which is expected for skills. It does write/produce local output files (script outputs), but does not try to modify other skills or system-wide configs.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install vectcut - After installation, invoke the skill by name or use
/vectcut - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.5.8
**Summary: Introduced a dedicated“视频配音混剪”混剪子链路，新增规则与脚本支持。**
- 新增视频配音混剪规则文档（rules/video-voiceover-remix.md）
- 增加混剪链路的 4.1 子链路“video-voiceover-remix”，包含固定七步流程
- 默认指定配音音色和文案字数估算方式，并提示可查官网获取音色
- 混剪二步重排支持调用内置脚本 scripts/remix_and_add_videos.py
- 新增相关脚本文件和 pyc 缓存文件以支持流程自动化
v1.5.7
- 新增 cut-koubo-template 模板口播编排链路支持，适用于固定风格和降低不确定性的场景
- 口播成片场景新增“模板稳定链”，严格校验模板参数，定制化输出
- 目录索引补充 cut-koubo-template 规则文档
- 完善模板链路的草稿生成、首帧处理与用户交互说明
v1.5.6
Version 1.5.6
- Added references/ai-image-model_capabilities.json to the project files.
- No changes to main logic or user-facing documentation.
v1.5.5
- Refactored script organization: moved all script files from nested directories to a flat top-level scripts/ directory.
- Removed legacy scripts in subfolders; corresponding new scripts added directly under scripts/.
- No changes to documented user scenarios or rules in SKILL.md.
v1.5.3
- Centralized and reorganized reference files for BGM, effect audio, model capabilities, and cover prompts into a single references directory.
- Removed duplicate and dispersed reference files from feature-specific subdirectories.
- No user-facing logic or behavior changes; update only affects internal file structure and data management.
v1.5.2
No changes detected in this version.
- No file or documentation updates found for version 1.5.2.
- Functionality and usage remain the same as the previous release.
v1.5.0
Major update: standardized skill聚合 and streamlined documentation for easier orchestration.
- 全面重写 SKILL.md，转向聚合编排式说明与场景/接口路由，聚焦用户视角与高频需求。
- 精简并移除大量 legacy 文件、用例、旧 endpoint 说明，仅保留规则、参考与 API search 功能。
- 新增或整理 20+ 规则文档与引用样例，覆盖主流剪辑能力、API路由与通用问题自查。
- 输出要求更清晰：链路执行摘要、失败追踪、必要字段结构化输出等。
- 明确环境变量检查与自动 API search 兜底逻辑，提升易用性与故障自愈能力。
v1.1.9
**Vectcut v1.1.9** — Major feature expansion
- 新增 file_manager、koubo、scrapt、subtitle_template 四大能力域，包含规则、端点、脚本、提示与样例，显著提升字幕模版、口播模版、爬虫解析及文件上传管理能力
- 新增 32 个文件，完善领域内参考文档（prompt、rules、reference）、curl 示例与执行脚本，丰富开发与验证工具链
- 优化 generate_ai_video、tts_generate、generate_ai_image 端点参数和输出格式，提升通用性和易用性
- 拆分/规范能力域文档索引，完善领域说明与枚举引用，提升可维护性
- 移除冗余枚举文件 references/enums/text_outro_anims copy.json
v1.1.6
vectcut-skill 1.1.6
- No code or documentation changes detected in this version.
- No updates were made to the SKILL.md or other files.
v1.1.5
- 增加了主页地址和 Openclaw 兼容元数据（homepage, emoji, requires, primaryEnv 字段）。
- 明确声明依赖的环境变量 VECTCUT_API_KEY。
- 其他内容未变，无代码或核心功能修改。
v1.1.3
**Major upgrade focused on modularity, scripted orchestration, and advanced autonomous video planning.**
- Re-architected skill for modular “能力域” (domain) structure, with rules, params, prompts, scripts, and examples per domain.
- Introduced autonomous video planning logic: AI can analyze content, plan script structure, and manage the full editing workflow.
- Added many demo scripts and modular prompt/rules/references directories; removed old hardcoded prompt/reference assets.
- Enhanced support for ASR, AI-generated media (image, video, voice), digital human, and keyframe editing, with clear execution order and output conventions.
- Streamlined environment configuration and interaction conventions for user clarity and scriptability.
v1.0.0
vectcut-skill 1.0.0 初始发布：
- 封装 curl 调用 HTTP API，自动化创建并编排剪映/CapCut 草稿、素材、字幕、特效和渲染任务。
- 支持一站式视频草稿创建、素材与时间线编辑、AI 生图/配音、智能抠像和批量生产等多种视频生产流程。
- 提供丰富接口：添加/编辑各类素材、拉取动画/特效/滤镜/字体等可用枚举、发起云渲染并实时查询渲染状态。
- 输出规范化结果，便于后续流程自动衔接（含 draft_id, draft_url, task_id）。
- 适合批量自动化视频制作和 AI 驱动的短视频生成场景。
Metadata
Frequently Asked Questions
What is VectCut: The first "To-Agent" video editing platform. Build automated video workflows with our stable cloud APIs in Coze, Dify, or OpenClaw.?
VectCut 全能剪辑聚合技能。用于剪口播、混剪、电商广告等场景，统一编排流光剪辑能力（字幕、音频、特效、画中画、抠像、AI补镜、AI配音、平台视频抓取、渲染导出）。当用户提出“做成片/自动剪辑/口播包装/广告剪辑/抓平台视频后再剪”等需求时，必须优先使用本技能。 It is an AI Agent Skill for Claude Code / OpenClaw, with 1328 downloads so far.
How do I install VectCut: The first "To-Agent" video editing platform. Build automated video workflows with our stable cloud APIs in Coze, Dify, or OpenClaw.?
Run "/install vectcut" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is VectCut: The first "To-Agent" video editing platform. Build automated video workflows with our stable cloud APIs in Coze, Dify, or OpenClaw. free?
Yes, VectCut: The first "To-Agent" video editing platform. Build automated video workflows with our stable cloud APIs in Coze, Dify, or OpenClaw. is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does VectCut: The first "To-Agent" video editing platform. Build automated video workflows with our stable cloud APIs in Coze, Dify, or OpenClaw. support?
VectCut: The first "To-Agent" video editing platform. Build automated video workflows with our stable cloud APIs in Coze, Dify, or OpenClaw. is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created VectCut: The first "To-Agent" video editing platform. Build automated video workflows with our stable cloud APIs in Coze, Dify, or OpenClaw.?
It is built and maintained by sun-guannan (@sun-guannan); the current version is v1.5.8.
More Skills