agentleFS
coltonjosephdean-rgbcoltonjosephdean-rgb/Hyperframes-colton.ai.dean/.claude/skills/setup/SKILL.md

setup

Skill19 starsChanged 5 months agoMITView on GitHub

First-run onboarding for a freshly cloned Hyperframes workspace. Use when CREATOR.md still contains the HYPERFRAMES_SETUP_INCOMPLETE marker, when the user says "/setup", "help me get set up", "let's get started", "I just cloned this", "set up Hyperframes", or otherwise arrives at a brand-new, un-configured clone. Walks the user through an environment check, a creator-profile interview, collecting their favorite reference reels (auto-studied via /study-creator), and brand colors — then removes the setup marker.

  • Reads credentials
  • Installs packages

We split this skill into the rules it gives an agent

12 rules. 0 are common, 5 are distinctive, and 7 are very specific to this project.

  • 0

    Common

    Most of the 170 edited versions we read keep these.

  • 5

    Distinctive

    This author's own choices. Worth a look.

  • 7

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    The most important step is Step 3 (reference reels): Before they make anything, you collect a few of their favorite videos so the workspace …

    line 12
  2. Very specific:

    If node is missing or <22 → point them to https://nodejs.org (LTS).

    line 42
  3. Very specific:

    If FFmpeg is missing → brew install ffmpeg (macOS), choco install ffmpeg (Windows), or the distro package manager (Linux).

    line 43
  4. Distinctive:

    If Chrome is missing → install the latest Google Chrome.

    line 44
  5. Very specific:

    If npx hyperframes isn't found after npm install → npx skills add heygen-com/hyperframes --yes then retry npx hyperframes doctor.

    line 45
  6. Distinctive:

    Don't block on a perfect environment — note anything missing, tell them how to fix it, and continue.

    line 47
  7. Very specific:

    Ask, ideally in one AskUserQuestion batch, the essentials.

    line 51
  8. Very specific:

    Write the answers into CREATOR.md by editing the matching sections.

    line 60
  9. Very specific:

    Run them one at a time and give a one-line progress note after each: [2/4] Studied @creator — signature move: <one line>.

    line 70
  10. Distinctive:

    If a link fails (private account, region lock, login wall), tell them and move on to the next — don't abort the whole step.

    line 71
  11. Very specific:

    If they genuinely have none right now, that's OK — tell them they can add references any time with /study-creator <url>, and the workspace …

    line 73
  12. Distinctive:

    Ask for their brand look (or offer to start neutral):

    line 77

“Common” is measured: kept by at least 6 in 10 of the 170 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Setup — Get a New Hyperframes User Started

This skill is the welcome mat. Someone just cloned this repo into Claude Code and said "clone this repo" / "help me get set up." Your job is to walk them from zero to ready-to-make-their-first-video, in a friendly one-pass conversation.

Tone: warm, concise, encouraging. They may be brand new to video editing AND to Claude Code. Don't dump walls of text — one short step at a time, confirm, move on.

The most important step is Step 3 (reference reels). Before they make anything, you collect a few of their favorite videos so the workspace learns the style they want you to edit in. Don't skip it.

When this skill fires

  • CREATOR.md contains <!-- HYPERFRAMES_SETUP_INCOMPLETE --> and the user sends their first message
  • The user says /setup, "help me get set up", "let's get started", "I just cloned this", "set up Hyperframes"
  • Any time the user wants to re-run onboarding or reset their profile

Before you start

Read CREATOR.md and PREFERENCES.md so you know the current state. If the setup marker is gone and the profile looks filled in, ask the user whether they want to re-run setup (and overwrite) or just edit a specific section — don't blow away a configured profile without asking.

The flow

Step 0 — Welcome

Greet them. One or two sentences. Set expectations:

👋 Welcome to Hyperframes! This is a workspace where I build short-form videos (TikTok / Reels / Shorts) for you in your style. Let's spend ~5 minutes getting set up so I know who you are and what you like. I'll go one step at a time.

Step 1 — Environment check

Confirm the machine can actually render video. Run these and report results plainly:

node --version          # need Node 22+
npm install             # installs the workspace dev deps (just Playwright)
npx hyperframes doctor  # checks Node, FFmpeg, Chrome
  • If node is missing or <22 → point them to https://nodejs.org (LTS).
  • If FFmpeg is missing → brew install ffmpeg (macOS), choco install ffmpeg (Windows), or the distro package manager (Linux).
  • If Chrome is missing → install the latest Google Chrome.
  • If npx hyperframes isn't found after npm install → npx skills add heygen-com/hyperframes --yes then retry npx hyperframes doctor.

Don't block on a perfect environment — note anything missing, tell them how to fix it, and continue. They can render once it's installed.

Step 2 — Creator interview

Ask, ideally in one AskUserQuestion batch, the essentials. Keep options concrete; let them free-type when needed:

  1. Name on-screen — what name (or brand name) appears in outros/lower-thirds?
  2. Handles — Instagram / TikTok / YouTube / X (whichever they have; "—" for none).
  3. Platform priority — TikTok-first / Reels-first / Shorts-first / cross-post all. (Drives default aspect ratio; default to 9:16 vertical 1080×1920 unless they say landscape.)
  4. Content niche — what their videos are about (AI, fitness, finance, cooking, comedy, etc.).
  5. On-camera mix — face-cam / faceless / both.
  6. Workflow — do they record + pre-edit their own speaking video (you build the visual layer on top, never cutting their audio), or build from scratch with you (motion graphics + TTS)?

Write the answers into CREATOR.md by editing the matching sections. Replace the _(placeholder)_ italics with their real values. Do not remove the setup marker yet — that happens at the end.

Step 3 — Reference reels (THE KEY STEP — don't skip)

This is what makes the workspace edit in their style. Say something like:

Now the fun part. Before we make anything, paste in 3–5 of your favorite videos — reels, TikToks, or Shorts whose editing style you'd love yours to feel like. They can be other creators you admire or your own best videos. I'll study each one frame-by-frame and learn its signature moves so your videos start out looking the way you want.

Collect the links (any mix of TikTok / Instagram Reel / YouTube Short / YouTube URLs). For each link, run the existing /study-creator <url> skill — it downloads, transcribes, extracts frames, reads every frame, and writes a style analysis into _reference/creator-library/<handle>/. At 3+ videos per creator it auto-rolls a style-distilled.md fingerprint.

  • Run them one at a time and give a one-line progress note after each: [2/4] Studied @creator — signature move: <one line>.
  • If a link fails (private account, region lock, login wall), tell them and move on to the next — don't abort the whole step.
  • Suggest at least 3 total. If they only have one creator they love, 3 videos from that one creator still triggers a distilled fingerprint.
  • If they genuinely have none right now, that's OK — tell them they can add references any time with /study-creator <url>, and the workspace will lean on MOTION_PHILOSOPHY.md defaults until then.

Step 4 — Brand colors

Ask for their brand look (or offer to start neutral):

  • Background color, accent/highlight color, text color, and a display font (or "use sensible defaults for now").

Fill assets/brand-tokens.css with their values. If they're not sure, set a clean dark default and tell them it'll evolve via /feedback:

:root {
  --brand-bg: #0a0a0f;
  --brand-surface: #1a1a23;
  --brand-accent: #3b82f6;
  --brand-text: #ffffff;
  --font-display: system-ui, sans-serif;
}

Step 5 — Optional API keys

Mention that nothing is required to start, but if any project uses OpenAI or ClickUp, they can copy .env.example to .env and add keys later. Don't block on this.

cp .env.example .env   # then fill in keys only if/when a project needs them

Step 6 — Finish

  1. Remove the setup marker. Edit CREATOR.md and delete the <!-- HYPERFRAMES_SETUP_INCOMPLETE --> line (and its companion comment lines). This flips the workspace into normal mode so the welcome doesn't fire again.
  2. Recap in 2–3 sentences: what you saved (profile, N reference creators studied, brand colors), and that PREFERENCES.md will fill in automatically as they use /feedback.
  3. Point them at the first build:

    You're all set! 🎬 To make your first video, say /make-a-video (I'll interview you and build it end-to-end) or /short-form-video if you've already got a talking-head clip to overlay. After each video, run /feedback so I keep getting closer to your taste.

Safety rules

  • Never remove the setup marker before the flow actually completes — if the user bails partway, leave it so onboarding resumes next time.
  • Don't overwrite a already-configured CREATOR.md/PREFERENCES.md without confirming.
  • Don't fabricate reference analyses — only /study-creator writes to _reference/creator-library/. If a link fails, say so.
  • Keep it conversational — one step at a time, confirm before moving on. This is someone's first impression of the workspace.

Download now (ZIP)

  • Other versions190 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 190

    1. anthropics/knowledge-work-plugins★ 34ksame name
    2. garrytan/gbrain★ 31ksame name
    3. qwibitai/nanoclaw★ 31ksame name
    4. gavrielc/nanoclaw★ 31ksame name
    5. nanocoai/nanoclaw★ 31ksame name
    6. Q00/ouroboros★ 6.3ksame name
    7. pytorch/executorch★ 5.1ksame name
    8. agenticnotetaking/arscontexta★ 3.5ksame name
    9. aws/agent-toolkit-for-aws★ 3.4ksame name
    10. spotify/portal-ai-plugins★ 2.8ksame name

    180 more not listed

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • Used alongsideProjects with it also use add-parallel, x-integration and 10 more.

    Other files people use alongside this one.

    Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
    • MicrosoftSkill

      add-parallel

      Also in 8 projects that use this

    • MicrosoftSkill

      x-integration

      X (Twitter) integration for NanoClaw. Post tweets, like, reply, retweet, and quote. Use for setup, testing, or troubleshooting X functionality. Triggers on "setup x", "x integration", "twitter", "post tweet", "tweet".

      Also in 8 projects that use this

    • juancruzmunozalbeloSkill

      add-telegram-swarm

      Add Agent Swarm (Teams) support to Telegram. Each subagent gets its own bot identity in the group. Requires Telegram channel to be set up first (use /add-telegram). Triggers on "agent swarm", "agent teams telegram", "telegram swarm", "bot pool".

      Also in 7 projects that use this

    • +9 more from microsoft
  • In this repository11 other files it gives its agents.

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.