← 返回 Skills 市场
496
总下载
2
收藏
1
当前安装
1
版本数
在 OpenClaw 中安装
/install video-to-text
功能描述
Use when the user wants to transcribe, caption, or get the text content of a video or audio file — e.g. "transcribe this video", "get the transcript", "what...
安全使用建议
Transcription: safe and local — the provided script extracts audio, runs the Whisper CLI, and writes .txt/.srt files next to your input file; it needs ffmpeg and the openai-whisper CLI (pip). Before installing: verify you trust the openai-whisper package and have the required system resources for the chosen Whisper model. Privacy: transcripts remain local unless you explicitly use the optional AI Edit flow. AI Edit: the SKILL.md shows an upload flow that sends your video to a remote API (SPARKI_API_BASE / agent-api-test.aicoding.live) and requires a SPARKI_API_KEY—do not set or share that key unless you accept uploading media to that third-party service. If you plan to use AI Edit, review any upload code and the external service's privacy/TOS and verify the SPARKI endpoint is legitimate before enabling the key.
功能分析
Type: OpenClaw Skill
Name: video-to-text
Version: 1.0.0
The `scripts/transcribe.sh` file contains a shell injection vulnerability. The `$INPUT` variable, which is derived directly from user input via the `SKILL.md` tool definition, is used unsanitized within the `ffmpeg -i "$INPUT"` command. This allows an attacker to inject arbitrary shell commands. Additionally, the `SKILL.md` describes an 'AI Edit' tool that uploads user-provided video/audio files to an external API (`agent-api-test.aicoding.live`), which, while stated functionality, involves significant data transfer to a third-party service.
能力评估
Purpose & Capability
The skill's name/description (video → text) match the declared requirements (ffmpeg and whisper) and the included transcribe.sh implements exactly that. Required binaries are appropriate and proportionate.
Instruction Scope
The main runtime instructions and scripts operate locally (extract audio, call whisper, write .txt and .srt next to the input). However SKILL.md also documents an optional 'AI Edit' workflow that uploads files to an external API and requires a SPARKI_API_KEY; that networked behavior is outside core transcription and only applies if you enable/use AI Edit.
Install Mechanism
This is instruction-only (no install spec). It recommends installing ffmpeg and pip-installing openai-whisper (whisper CLI). No remote archive downloads or unusual install locations are used by the skill itself.
Credentials
The skill itself requires no environment variables. The optional AI Edit feature requires a SPARKI_API_KEY (external service) — this is not required for transcription but will be needed if you use AI Edit, which uploads assets to a third-party endpoint.
Persistence & Privilege
always:false and user-invocable:true. The skill does not request elevated persistence or modify other skills' configs. The SKILL.md suggests using openclaw config to store SPARKI_API_KEY if you opt in, which is normal for enabling external integrations.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install video-to-text - 安装完成后，直接呼叫该 Skill 的名称或使用
/video-to-text触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
- Major update: Skill refocused from general video processing to dedicated video/audio transcription.
- Renamed skill to "video-to-text" with new description and clarified target use cases (transcription, subtitles, speech-to-text).
- Removed platform/quality/workflow/command reference files and general editing scope.
- Added local transcription workflow using Whisper (no API key needed) with clear installation and usage instructions.
- Skill now outputs both plain text and SRT subtitle files.
- Provided guidance to escalate to AI Edit tool for advanced, transcript-driven video editing tasks.
元数据
常见问题
Video Transcribe 是什么？
Use when the user wants to transcribe, caption, or get the text content of a video or audio file — e.g. "transcribe this video", "get the transcript", "what... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 496 次。
如何安装 Video Transcribe？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install video-to-text」即可一键安装，无需额外配置。
Video Transcribe 是免费的吗？
是的，Video Transcribe 完全免费（开源免费），可自由下载、安装和使用。
Video Transcribe 支持哪些平台？
Video Transcribe 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（darwin, linux）。
谁开发了 Video Transcribe？
由 BoShen（@symbolk）开发并维护，当前版本 v1.0.0。
推荐 Skills