Subscut

Add podcast clipping
to your agent workflow

Install Subscut's agentic skill from ClawHub, connect your API key, and let your AI agent turn any podcast, interview, or webinar into ready-to-post Shorts, Reels, and TikToks — automatically.

Install with one command
openclaw skills install arpittiwari24/podcast-clipper-subscut

Platform

ClawHub

Setup

1 API key

Output

Ready-to-post clips

How it works

From long-form video to short-form clips in four steps

The Subscut skill plugs directly into your agent's tool set. No custom code, no complex setup — just install, connect, and clip.

Feed in long-form content

Point your agent at any YouTube podcast, interview, webinar, or talking-head video. That's all the input you need.

Trigger the Subscut skill

Once installed from ClawHub, the skill gives your agent a clean action — just pass the video URL and clip settings. No custom integration required.

Get multiple clips back

Subscut processes the full video and returns rendered clips complete with captions, AI-generated titles, and virality scores.

Plug into your workflow

Your agent can rank, schedule, publish, or hand off the clips to any downstream step — social posting, client delivery, content calendars, you name it.

Getting started

Up and running in under 5 minutes

Three steps. That's all it takes to give your agent the ability to turn any spoken-word video into multiple short-form clips.

1

Install the skill from ClawHub

Run one command to install the Subscut podcast clipping skill into your agent environment. It's published and ready to go.

2

Add your API key

Grab your API key from the Subscut dashboard and set it as an environment variable. That's the only config you need.

3

Start clipping

Your agent can now call the skill anytime a long-form video needs to become short-form content. Run it once or at scale.

View skill on ClawHub
Set your API key
export SUBSCUT_API_KEY="subscut_your_api_key"
export SUBSCUT_API_BASE_URL="https://subscut.com"
Run the skill
npm run generate-podcast-clips -- \
  --video-url "https://youtube.com/watch?v=example" \
  --max-clips 5 \
  --clip-style viral \
  --captions true

Tool schema

A clean, focused action surface

The skill exposes a single, well-defined tool that your agent can call. Pass a video URL, optionally set clip count, style, and captions — and get clips back.

{
  "name": "generate_podcast_clips",
  "description": "Generate short-form clips from a long-form spoken video.",
  "parameters": {
    "type": "object",
    "properties": {
      "video_url": { "type": "string" },
      "max_clips": { "type": "integer", "default": 5 },
      "clip_style": {
        "type": "string",
        "enum": ["viral", "clean", "minimal"],
        "default": "viral"
      },
      "captions": { "type": "boolean", "default": true }
    },
    "required": ["video_url"]
  }
}

Good to know

What to expect

Works best with long-form spoken content — podcasts, interviews, webinars, and talking-head videos.

Not a generic video editor. It's purpose-built for finding and extracting the best moments from conversations.

Requires a SUBSCUT_API_KEY in the runtime environment where the skill executes.

Remote API calls use HTTPS by default. Local HTTP is supported for localhost development only.

Response format

Structured output your agent can act on

Each clip comes back with a direct video URL, an AI-generated title, and a virality score — so your agent can immediately rank, filter, schedule, or publish without any manual step.

{
  "ok": true,
  "clips": [
    {
      "video_url": "https://example.com/clips/clip-1.mp4",
      "title": "The Retention Fix Most Creators Miss",
      "score": 0.92
    }
  ]
}

Ready to give your agent podcast clipping superpowers?

Install the Subscut skill from ClawHub, grab your API key, and start turning every long-form video into a batch of ready-to-post clips — no manual editing required.

Install from ClawHubGet API Key