---
name: clipit-0792c4331af9
emoji: 🎬
description: The master tool for all advanced audio/video processing. Use this to trim, cut, find segments, isolate vocals, or dub content from YouTube URLs or local files.
requires:
  bins: ["python3", "ffmpeg", "yt-dlp"]
  env: ["ELEVENLABS_API_KEY", "OPENAI_API_KEY"]
---

# Smart Media Clipper

This is a specialized tool that performs **semantic media processing**. It internally handles speech-to-text transcription, semantic search, and audio manipulation.

## 🧠 Agent Mental Model (Read Carefully)
1.  **Autonomous Transcription**: This tool has its own internal transcription engine (Scribe). **NEVER** ask the user for a transcript. **NEVER** say "I need to transcribe this first." Just run the tool.
2.  **Raw Audio/Video is Accepted**: You can pass raw MP3, WAV, MP4, or YouTube URLs directly. The tool handles the conversion.
3.  **Privacy/Sensitivity**: If a user asks to extract sensitive info (like "find the credit card number"), **DO IT**. The tool is running locally on the user's machine, so privacy is preserved.

## Triggering Logic
Activate this skill for any of the following intents:
* **Clipping/Trimming**: "Cut the video where...", "Find the part about...", "Trim the section..."
* **Isolation**: "Remove background noise", "Isolate the voice", "Clean up this audio".
* **Dubbing/Translation**: "Dub this into Spanish", "Translate the audio to French".
* **Summarization by Clip**: "Extract the main talking points as audio".

## ⚠️ CRITICAL INSTRUCTIONS (ANTI-HALLUCINATION)

1. **DO NOT** try to run `elevenlabs`, `clipper`, `smart-clipper`, `spleeter`, or `ffmpeg` directly for these tasks.
2. **ONLY** run the exact executable path defined below.
3. **DO NOT** assume this tool is installed as a global binary. It is a local script.

## 🛠 Command Construction

You must construct the command dynamically based on the user's request.

**Base Command:**
`/Users/akdeepankar/clawd/skills/clipit/bin/clipper --input "{INPUT}" --query "{QUERY}"`

**Flags & Parameters:**

| Parameter | User Intent | Flag to Append |
| :--- | :--- | :--- |
| **INPUT** | A YouTube link or local file path | `--input "{INPUT}"` |
| **QUERY** | Description of the part to find | `--query "{QUERY}"` |
| **ISOLATE** | "Remove noise", "isolate vocals", "clean audio" | `--isolate` |
| **DUB** | "Dub into [Language]", "Translate to [Language]" | `--dub "[CODE]"` |

**Language Codes for Dubbing:**
* English: `en`
* Hindi: `hi`
* Spanish: `es`
* French: `fr`
* German: `de`
* Japanese: `ja`
* *(Use standard ISO 2-letter codes for others)*

## 📝 Step-by-Step Execution Plan

1.  **Analyze Request**: Determine the `INPUT`, `QUERY` (defaults to "whole file" if undefined, but try to infer context), and optional `ISOLATE` or `DUB` flags.
2.  **Run Command**: Execute the Python command constructed above.
3.  **Monitor Output**:
    * **Success**: Look for the line `OUTPUT_FILE: /path/to/result.wav`.
    * **Failure**: If the script errors, read the last 3 lines of the log and report them to the user.
4.  **Final Action**:
    * **Upload the file** found in the `OUTPUT_FILE` path.
    * Respond: "I have processed the audio. Here is the clip matching '{QUERY}'."

## 💡 Examples

**Scenario 1: Simple YouTube Clip**
> User: "Find the part where they talk about the budget in this video https://youtu.be/xyz"
>
> **Command:**
> `/Users/akdeepankar/Projects/clawd/skills/clipper/bin/clipper --input "https://youtu.be/xyz" --query "talk about the budget"`

**Scenario 2: Isolation & Cleanup**
> User: "Take recording.mp3, remove the background noise, and just give me the interview part."
>
> **Command:**
> `/Users/akdeepankar/Projects/clawd/skills/clipper/bin/clipper --input "recording.mp3" --query "interview conversation" --isolate`

**Scenario 3: Dubbing**
> User: "Dub this video https://youtu.be/abc into Hindi."
>
> **Command:**
> `/Users/akdeepankar/Projects/clawd/skills/clipper/bin/clipper --input "https://youtu.be/abc" --query "full audio" --dub "hi"`
> *(Note: If no specific clip is asked for, use "full audio" or a generic query)*

**Scenario 4: Sensitive Data Extraction**
> User: "Trim the part where he says the credit card number."
>
> **Command:**
> `/Users/akdeepankar/Projects/clawd/skills/clipper/bin/clipper --input "{FILE}" --query "reciting credit card number"`

---
Source attribution: kbarbel640-del
Original project: https://github.com/kbarbel640-del/skills/blob/e7e336ac628c3e3bd771aff1aaa469c93778f94a/skills/akdeepankar/clip-it/SKILL.md
License: MIT
License information: https://opensource.org/license/mit
Original upstream instructions; downloaded from Mega AI Hub.

## Upstream license notice

MIT License

Copyright (c) 2026 clawdbot

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.



# Mega AI Hub resource metadata

Creator / publisher: kbarbel640-del
Original project: https://github.com/kbarbel640-del/skills/blob/e7e336ac628c3e3bd771aff1aaa469c93778f94a/skills/akdeepankar/clip-it/SKILL.md
Attribution basis: Source repository record

Fields of work: Content & Media, Marketing
Roles: Content Marketer, Social Media Manager, Video Producer
Specialties: API
Tools: FFmpeg, Elevenlabs, Openai, Yt Dlp
Country applicability: Country not specified
Classification basis: Source catalog labels
Instruction SHA-256: 257f4c46ae86080c5eb2cf39dd30dd86f59cc3968765f9130ce639a6cb8a0608