Video resume
A stack of skills for your AI agent, made for people starting from zero. 103 skills and 149 project briefs — load one into Codex, Claude Code, Copilot, Gemini, or Cursor and do the thing.
npx -y skills add protosphinx/sphinxstack --skill video-resumeAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 17 days oldThe repository was created 17 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Record a short video introduction for an application that explicitly requests or accepts one. Covers the person's own script, recording, captions, privacy, and controlled sharing. Use when an application asks for a video introduction or the person deliberately chooses a video accepted by the recipient.
SKILL.md
4.2 KB, 931 tokens by cl100k_base, as published. Nobody here has run it
video-resume
Make a 60-second video introduction with someone — them, on camera, saying who they are and what they have done. Use this only when the application explicitly requests or accepts a video, or when the person has confirmed that the recipient welcomes one. An unsolicited video can expose personal information and introduce bias; do not add it to every resume by default. The session ends with a reviewed video and a controlled share link.
Privacy and consent
- Do not state an exact age, home address, school schedule, identity number, or other private detail. Use only the contact information already approved for the application.
- Use an unlisted or access-controlled link. Explain that anyone with an unlisted link may still copy or forward it.
- The person decides whether the benefit is worth the exposure. If the application does not accept video, stop and use the written resume.
Script: their words, their facts
- Everything said must be true and match the written resume (build it first with the build-resume skill if it does not exist).
- Structure, four beats, ~60 seconds total:
- Who: name, and the one-line version of what they do or study.
- Proof: the strongest one or two real things — a project with a link, a job, a result. Specifics carry this beat; "I built X, it does Y, here is where it lives."
- Want: what they are looking for, concretely.
- Close: name again, "thanks", done. No catchphrases.
- Write it by interview: ask, they answer out loud, you transcribe and tighten. Read the draft back — if any sentence sounds like a LinkedIn post instead of like them talking, rewrite it in what they actually said. Word count check: ~140–150 words is a minute.
- They rehearse from beats, not from the full text. Memorized scripts read as recited; beats keep it a person talking.
Setup
Phone camera, and it is enough:
- Horizontal, at eye level (stack of books works), arm-and-a-half away. Lens at eye height matters more than any other setting.
- Light: face a window, daytime. Never a window behind them.
- Background: a plain wall or a tidy real corner of their room. Real beats staged; messy beats neither.
- Audio: quiet room, or wired earbuds as the mic — run the record-clean-audio checks. Speech must remain clear on ordinary speakers.
- Clothes: whatever they would wear to an interview for this target. A 5-second test clip to check framing and sound first.
Takes and edit
- Full run-throughs only, no stopping mid-take: fumbles included, straight to the end, then again. Take 1 is a warm-up by definition. Most people land it between takes 3 and 6; past 8, break, because tired takes get stiffer.
- Pick the take where they sound most like themselves — not the one with zero fumbles. A small stumble and a recovery reads human.
- Edit lightly in CapCut or DaVinci free: trim to start on their first word and end after the close, level the audio, done. No music, no titles, no transitions. Optionally cut between the best two takes if the framing matches.
- Watch it together once at phone size and once with sound off — posture and framing should hold up mute.
Ship
- Export 1080p MP4.
- Upload to YouTube as unlisted — findable by link only, not searchable, not on their channel page for subscribers to trip over. Title it "Firstname Lastname — Introduction".
- Caption it (captions-and-subtitles skill) — recruiters watch muted at desks.
- Put the link only in the application field or message where video was requested or confirmed as welcome. Test it from a logged-out browser.
Done
- The recipient accepts video and the person made an informed sharing choice
- Link works logged out, captions are correct, and every claim matches the written application
- No unnecessary personal information appears in speech, background, title, description, or account profile
Gives 0 of the 12 instructions most video audio skills give in 931 tokens
Counted across 621 of the 795 authors here whose files we hold, read 2026-08-06
- read individual rule files for detailed explanationsin 21 of 621, across 9 files
- Use WAV PCM 16kHz mono audio formatin 13 of 621, across 4 files
- render final videoin 13 of 621, across 6 files
- use this skill when dealing with Remotion codein 11 of 621, across 4 files
- save generated audio to a WAV filein 11 of 621, across 4 files
- handle conversion errors gracefullyin 10 of 621, across 6 files
- add captions to videos alwaysin 10 of 621, across 4 files
- generate music from text descriptions using MusicGenin 9 of 621, across 2 files
- do not skip pipeline layersin 9 of 621, across 3 files
- do not make one tool do everythingin 9 of 621, across 3 files
- never ask the user to paste their full API keyin 9 of 621, across 3 files
- use azure document intelligence for complex pdfsin 9 of 621, across 4 files
Said here and by no other author read
- use only when video is explicitly accepted
- build the written resume first
- exclude private details like age and address
- use approved contact information only
- match every claim to the written resume
- write the script by interview
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.