Features

Everything your agent needs to get things done.

We extract as much information as possible from every clip. Anything that crosses your screen becomes available to your agent. Everything’s searchable and actionable.

Other recorders store video files. ClipCabinet stores answers.

ClipCabinet turns your clip into a transcript, captioned frames, and every URL, error, code snippet, and ticket ID that crossed the screen, all searchable and readable by your agent. Ask it “what was the error at 2:14?” and it quotes the actual error.

01 Capture

One click, any tab.

The recording extension quickly records your active tab, system, and mic audio. No fuss, no setup, just click, talk, and go.

  • Quick access to recordings from your toolbar.
  • Automatically pulls out artifacts from the recording source.
  • Floating HUD lets you pause, finish, or discard from any page.
  • Screenshot mode for quick comments and feedback.
02 Pipeline

Stop recording. Get a structured clip back.

The moment you hit stop, ClipCabinet starts turning your clip into a fully indexed record your agent can query.

  • Word-level transcript with timestamps.
  • Captioned keyframes describe what's actually on the screen.
  • Every URL, error, code snippet, file path, and ticket ID pulled out as structured data.
  • One-paragraph summary your agent can quote in a heartbeat.
  • Vector embeddings on every layer. Semantic search works across your whole library.
03 Native MCP

Seven tools your agent calls without you.

We have all the tools so your agent can work autonomously, with instructions generated automatically on the fly. ClipCabinet speaks Model Context Protocol natively, so Claude Code, Cursor, and Claude Desktop already know how to query it. Paste the config, mint a token, and your agent does the rest.

list_recordings

Recent clips by date or kind (video / screenshot).

search_recordings

Semantic search across the whole library.

get_recording

Full details: transcript, frames, artifacts, signed media URLs.

search_within_recording

Find moments inside one clip.

get_frame

A captioned keyframe at a given timestamp.

get_audio_clip

An audio slice between start and end timestamps.

list_artifacts

Every URL, error, or ticket ID across clips.

.mcp.json
{
  "mcpServers": {
    "clipcabinet": {
      "type": "http",
      "url": "https://mcp.clipcabinet.com/mcp",
      "headers": {
        "Authorization": "Bearer kg_xxxxxxxxxxxxxxxxxxxx"
      }
    }
  }
}

Read more about MCP in our docs.

04 Click-to-Comment

Comment at a pixel, not a timestamp.

Almost every tool out there only allows comments that are time-based. ClipCabinet's comments live on an X-Y axis, not just a timestamp. Click anywhere on a paused frame or a screenshot to drop a pin at that exact spot.

  • Pin lives at the exact spot you clicked, not just a timestamp.
  • Convenient color-coded pins. Keep things organized.
  • Share your clips and allow others to add comments and collaborate.
05 Markdown Twin

Every clip has a Markdown URL.

Paste a URL and get the raw markdown. Great for API endpoints and direct data sharing with your agents for any of your team pipelines.

  • GET /recordings/[id].md or /s/[token].md returns the full Markdown payload.
  • Share links unfurl in Slack, Discord, and X with a video preview.
  • Webhooks fire on recording.completed, recording.failed, comment.added.
  • Public share pages render without an account; agents read them too.
06 Private & Portable

Private by default.

Every clip is private until you decide to share it. Share links are public, opt-in, and revocable. Nothing is locked in. Export any clip as markdown and send it where it's needed.

  • Clips are private by default; you choose what to share.
  • Public share links are opt-in and revocable anytime.
  • Export any clip as Markdown: transcript, frames, and artifacts.
  • API tokens are hashed at rest and revocable from settings.

Get your agent on the same page.

Start for free, no card needed.