paste this link into your ai. it will know what to do.
Package and share the videos made for kt (Claude Motion films, rendered MP4/GIFs, ElevenLabs/Figma exports, edits) with someone else, in one clean, permission-checked handoff, and log what went to whom.
share-videos
kt often needs every video made in a session (or for a project) handed to someone else: Daksh, Nischay, Sahil (Jonna), Shaurya, a client, an investor, a creator. This skill turns a pile of renders into one handoff the recipient can open in one tap.
0. Standing rule
- Every time a video is finished, add it to the session's video list (step 1) without being asked, so sharing is one step later.
- Nothing is sent, and no sharing permission is changed, without kt's yes in chat. Drafting is free; sending is not.
1. Build the video list
For every video made or touched in this session/project, record:
| # | title | project | what it is (one line) | version | length | where it lives (artifact link / Drive / file) |
|---|
Rules:
- Final versions only. Drop drafts, test renders and superseded versions unless kt asks for them.
- If two versions exist, name which one is final and confirm with kt.
- File names:
<project>_<title>_v<n>_<yyyy-mm-dd>.mp4(lowercase, hyphens inside words). Rename before sharing.
2. Settle the three unknowns (ask once, in one message, only what's missing)
- Who — name and channel (email, WhatsApp, Slack, Instagram DM).
- Access — view only (default), comment, or edit/download originals.
- Ask — what they should do: just watch, give feedback (by when), post it, or approve it.
If the recipient is a known team member, use what innernet already knows about them and skip the question.
3. Pick the route per video
- Claude Motion / claude.ai artifact: it is private until kt shares it. Give kt the card and tell him to use Share on it (or Export to save an MP4/GIF if the recipient needs a file). Claude does not change artifact sharing itself.
- A video file (MP4/GIF/MOV): upload to Google Drive under
Shared videos/<project>/<yyyy-mm-dd> - <recipient>/, then share that folder with the recipient's email at the chosen access level (needs kt's yes). One folder link beats ten file links. - Already on Drive / YouTube (unlisted) / Vimeo / Fathom: reuse the existing link; check its permission instead of re-uploading.
- Email: never attach large videos. Gmail caps attachments around 25 MB, so send a Drive/artifact link instead.
- Many videos (4+): also offer a one-page reel (an artifact) listing every video with title, one line and link, so the recipient gets a single link.
4. Check before it goes
- Every link opens for the recipient (right account / anyone-with-link), not just for kt.
- Correct, final version behind every link; no internal notes, watermarks-for-review or private drafts.
- Titles match the file names and the message.
- Nothing confidential to another venture is in the folder (innernet, Placerra, Frame & Kin, Howl & Mane, Prowess, J.K Enterprises stay separate unless kt says otherwise).
5. Draft the handoff message (kt's voice, short)
Structure:
- One line of context (what these are, why they're coming to you).
- The list:
title — one line — link (length). - The ask and the deadline.
Adapt to channel: email gets a subject (<project>: <n> videos for <purpose>); WhatsApp/Slack stays to a few lines. Show kt the draft; send only on his yes. Report back what was sent, to whom, and where.
6. Log it in innernet
After sending, save to the project the videos belong to (innernet_save_context, dimension shared_videos, appending): date, recipient, channel, access level, the list of titles + links, the ask, and the deadline. When feedback comes back, add it under the same entry.
Defaults (when kt isn't around to answer)
View-only access, Drive folder route for files, a drafted message waiting for approval, nothing sent.
keep it where your ai can reach it.
innernet is memory your ai tools read live: projects, decisions, skills, the things you'd otherwise re-explain every session. save this skill to yours, or publish your own as a link like this.