Written from the video, not about it
The title, description and tags come from the real transcript, so they match what viewers (and the algorithm) will actually hear — no generic keyword salad.
"Get this ready to upload" — title, chapters, subtitles, thumbnail
The video is done; the packaging is the chore. Claude runs the full kit from one file — SEO metadata written from the actual transcript, timestamped chapters, an uploadable .srt, and a thumbnail anchored to a real frame — and hands you one copy-paste block for YouTube Studio.
You type
Prep my latest video for YouTube — metadata, chapters, subtitles, thumbnail.
…and Claude drives the whole pipeline below on your machine.
Why this beats a blank script box
The whole checklist in one pass
Metadata, chapters, subtitles and thumbnail are four separate tools you'd normally open. One message runs them all and returns one upload-ready block.
You stay the publisher
ViralMint deliberately doesn't auto-upload. You get files and copy-paste text — full control over what goes live on your channel, with none of the OAuth risk.
How Claude builds a YouTube upload kit
-
Metadata from the transcript
Whisper transcribes the video locally; the AI writes a <70-char title, description, 5-10 tags and a TikTok caption from what's actually said — not from a guessed topic.
-
Chapters with timestamps
For long-form, the transcript's segments are grouped into 5-15 topical chapters, formatted ready to paste into the description.
-
An uploadable .srt
The same local transcription exports as a clean SRT cue file for YouTube Studio's subtitle upload — better than auto-captions, no burn-in.
-
A thumbnail from a real frame
Claude previews six frames, picks (or asks you to pick) the strongest moment, and composites the hook text over that actual frame — or generates AI variations if you want them.
MCP tools it calls:
tool_metadatatool_auto_chapterstool_subtitlespreview_framesthumbnail_from_frame Connect it in 4 steps
-
Install ViralMint
Download the free, open-source desktop app (macOS / Windows / Linux) and open it. It bundles the whole pipeline — yt-dlp, Whisper, FFmpeg — and the MCP server.
-
Open the in-app MCP page
It shows your loopback endpoint and a bearer token, with a ready-to-paste Claude Code command and a JSON config snippet for Cursor / Claude Desktop.
-
Add the server to your agent
Paste the command. Your client loads ViralMint's tools, the built-in workflow recipes (this skill is one) and read-only resources like your balance and recent videos.
-
Ask for the outcome
Type the prompt above. Claude runs the recipe end-to-end against the local app and hands you the finished mp4 — no clicking through the UI.
ViralMint skills vs typical agent video tools
| What matters | Typical paid agent tools | ViralMint |
|---|---|---|
| Agent tools / skills | Behind a paid tier | Free & open-source (AGPL) |
| Where it runs | Their cloud | Your machine (loopback MCP) |
| Your footage & data | Uploaded to their servers | Stays local — only AI calls leave |
| Pricing | Monthly subscription | Prepaid, pay-per-render at ~cost |
| Output watermark | On free tiers | Never |
| Script quality | Prompt → generic script | Grounded in real breakouts + your data |
highlighted column = clearer fit.
FAQ
Frequently asked
What exactly do I get back?
A YouTube title under 70 characters, a description with chapters pasted in, 5-10 tags, a TikTok caption with hashtags, the .srt file's path, and a thumbnail image — everything YouTube Studio asks for, in one block.
Do chapters work on Shorts?
Chapters are for long-form (they need multiple topical segments), so Claude skips that step for Shorts and focuses on metadata + thumbnail instead.
Is the thumbnail AI-generated or from my video?
Your choice. The default anchors to a real frame — Claude previews candidates and composites your hook text over the one you pick. AI-generated variations (faces kept recognizable) are available on request.
Can it prep several videos at once?
Yes — point it at your recent videos and it runs the kit per video, keeping each one's results grouped so nothing gets mixed up.
Make {skill.name.toLowerCase()} from a chat
Install the free ViralMint desktop app, connect it to Claude Code or Cursor, and run this skill — no subscription, no watermark, your footage never leaves your machine.