Papercut Hook logo

Papercut Hook

Community
0xS4N4TG
papercut-hook

Write or fix the hook of a papercut reel. Produces the SAY / WRITE / SHOW / ALIGN block, five variants scored with the hook check, the frame-0 cover title and the motion out of the hook. Use for "write a hook for...", or when a reel's first three seconds lose people.

Overview

Publisher0xS4N4TG
Repositorypapercut-explainer
Skill namepapercut-hook
Stars
106
Forks
10
Bundled files
1
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 1 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by 0xS4N4TG on GitHub. Read the source before you install it.

Installation

Install the Papercut Hook AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/0xS4N4TG/papercut-explainer/tree/main/skills/papercut-hook
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/0xS4N4TG/papercut-explainer.git /tmp/papercut-explainer
mkdir -p .claude/skills
cp -r /tmp/papercut-explainer/skills/papercut-hook .claude/skills/papercut-hook
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Papercut Hook in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Papercut Hook on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Papercut Hook is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Papercut hook

The hook is the most important step. A reel's hook is three separate, explicit layers, and none may be skipped or left implied. Write all of them before anything else in the script.

Preflight

  • Install: Minimal is enough. No library needed.
  • Read the BRIEF at the top of episodes/NN-slug/research/facts.md (topic, format, audience, CTA) and the facts list, so every claim can be checked. If there is no BRIEF, run papercut-strategy first.
  • Read SERIES.md for the audience level and regions.

The HOOK block

HOOK
SAY:   <the first spoken sentence>
WRITE: <the frame-0 title, 2-5 words, ideally a number; the episode.json "hook">
SHOW:  <what frame 0 and the first 2 s show: the mid-action moment, the camera move>
ALIGN: <the one idea all three point at>

No scene gets built until all four lines are filled and pass the alignment test.

LayerRule
SHOW, the first frameThe most striking moment, already mid-action. No establishing shot, no fade-in, no empty room. It must make sense muted.
WRITE, the on-screen titleThe cover title (2-5 words, ideally a number), fully visible on frame 0 for about 2-3 s. It carries the other half of the gap, never the same words as the voice.
SAY, the spoken lineOpens the gap within about 10 words. Names who it's for. The stake or number comes early. Never answers itself. No anonymous news framing ("Someone just released..."); a named, credible source is fine.

Skip the "stated as news" opening from beat 1 of the spine ("A solo dev just open-sourced..."). Open the gap and the stake instead, and name a big brand the viewer already knows in the first sentence.

Alignment test

What you show, what you write and what you say all point at one idea. They differ in words, never in subject. If the voice says "your screenshots leaked" while frame 0 is a calm desk and the title is about stars, each layer opens a different loop and none lands. Cover any two layers: the remaining one should still hint at the same idea.

Variants and the hook check

  1. Write 5 variants from at least three of the seven patterns in references/hook-patterns.md (borrowed authority + giveaway, name the viewer then a number, one object one number one twist, price shock, pain, verdict or conflict, own authority).
  2. Score each with one point per yes:
    1. Does the spoken line open a gap (a question or tension the viewer needs closed) within about 10 words?
    2. Does it say who it's for, or what is at stake?
    3. Does it carry one concrete number, object or twist instead of a vague claim?
    4. Does it leave the answer for later?
    5. Does frame 0 plus the title work with the sound off?
    6. Is every claim true to the facts sheet?
  3. Pick the variant with the most points. Drop any variant that fails question 6.

The frame-0 title (episode.json "hook")

Every reel opens on its hook. The hook key in episode.json puts the cover title on screen from frame 0 (build_vo.py writes it into plan.json) and holds the banner back until the title leaves.

"hook": {"lines": [["AI", "LEAKED"], ["13,000+"]], "hot": ["13,000+"], "until": "yes", "size": 108, "top": 220}
  • lines: the WRITE text as torn tags, one array of words per line. hot: the words that get the yellow tag.
  • The title leaves at the onset of the until word (["yes", 2] for its 2nd occurrence), or after "hold" seconds (default 2.4).
  • Title stays 2-5 words and shares no words with SAY.

The cut out of the hook carries motion

A hook that ends on a held pose feels abrupt and drops the attention it just earned.

  • Give the hook scene's last ~10 frames an accelerating push (a tiny pull back first).
  • Scene 2 opens on an impact at its focal prop, then the camera eases back to the set. For example, the proof card slams onto the board with a stamp, the camera starts at 1.17x, and a character scurries in from off-frame.
  • End the title on the cut with "hook": {"hold": <S01 seconds>} so it lifts away during the push instead of covering scene 2's first frames.

Gate

  • SHOW is mid-action on frame 0 and reads muted.
  • WRITE is on frame 0 and isn't the spoken words.
  • SAY opens the gap within 10 words.
  • All three pass the alignment test, and ALIGN names the one idea.
  • The winning variant scored best and passed question 6.

Hand-off

Write the HOOK block at the top of episodes/NN-slug/papercut/STORYBOARD.md (create the file if it isn't there). Note the hook's key word or number; the script's last line repeats it. Next is papercut-script, which also puts SAY into line l1_hook of narration/script.json.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Papercut Hook AI skill do?

Write or fix the hook of a papercut reel. Produces the SAY / WRITE / SHOW / ALIGN block, five variants scored with the hook check, the frame-0 cover title and the motion out of the hook. Use for "write a hook for...", or when a reel's first three seconds lose people.

Why use Papercut Hook on TypingMind?

Because you install it once and use it with any model. Papercut Hook is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Papercut Hook in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/0xS4N4TG/papercut-explainer/tree/main/skills/papercut-hook. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Papercut Hook?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Papercut Hook?

As many as you like. As long as a model supports skills, you can use Papercut Hook with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Papercut Hook AI skill free?

Yes. It is published on GitHub by 0xS4N4TG under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇