Papercut Script logo

Papercut Script

Community
0xS4N4TG
papercut-script

Turn a link into a papercut reel script. Researches a sourced facts list, writes the 75-95 word script on the 10-beat spine, a banner fact track, a staging table with one set per banner, and the humour pass. Use for "write the script", "research this repo for a reel".

Overview

Publisher0xS4N4TG
Repositorypapercut-explainer
Skill namepapercut-script
Stars
106
Forks
10
Bundled files
2
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.

  • 2 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 Script 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-script
  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-script .claude/skills/papercut-script
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Papercut Script 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 Script 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 Script 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 script

Research, script, banner track, staging and gags are written together, because each one feeds the others. The full system, with the phrase bank and two staged examples, is references/storytelling.md. Gag and format banks: references/comedy.md.

Every beat runs three tracks that do different jobs:

  • Voiceover (shown as caption tags): tells the story, casual and second person, changes every sentence.
  • Banner: adds the hard fact, number or honest caveat the voiceover skips, changes every 1.8-3 s.
  • Set: literally illustrates a concrete noun from the voiceover, with the product UI doing the demo, changes with the banner.

Preflight

  • Install: Minimal is enough. Resolve $PAPERCUT_LIB only if you want the backdrop sheet (library/papercut/public/bg/backdrops/_sheet.jpg) when choosing sets. Each image costs about 0.7-1.5k tokens, so open only what the step needs.
  • Needs the BRIEF at the top of episodes/NN-slug/research/facts.md and the HOOK block at the top of episodes/NN-slug/papercut/STORYBOARD.md. Missing one: run papercut-strategy or papercut-hook first.
  • Research needs web fetch or search, and any browser tool for screenshots. Without them, ask the user for screenshots and facts.

1. Research and rough UI

  • Collect facts from the source with a source for each: name, maker, numbers, price, requirements, limitations, competitor numbers. Mark anything uncertain.
  • Grab the key screens (2-4), real labels and button text, the accent colour and hero images. The video rebuilds a rough version of the UI as papercut props, not a screen recording (references/storytelling.md section 5).
  • Confirm the role framing from the BRIEF. Never invent product behaviour. See the role-framing note in ../papercut-explainer/references/frame-breakdown.md.
  • Write the list to research/facts.md under the BRIEF. Every number in the script, banners and UI must trace back to it.

2. Pick the mode

ModeWhen
Open-source dropA free or local alternative to a paid product
News launch / worth-itA new paid tier, model or feature from a big company
Feature unlock / how-toA hidden setting or command that changes how a tool works

Hook formulas and the middle of each mode: references/storytelling.md section 6. Apply the hook update: open the gap and the stake, not the news.

3. Script on the spine

Write 75-95 words for 22-30 s, about 3.5-4 words a second, inside the chosen format, with a new set every 1.5-3 s. The 10 beats:

  1. Hook (0-3 s), already fixed by papercut-hook
  2. Proof or tension
  3. Reframe ("So instead of..., you can...")
  4. Name reveal ("It's called...", never before the hook)
  5. How it works, one concrete demo
  6. Escalation ("And X is just the start, because..." plus a stack of exactly three)
  7. Peak ("And get this:" plus the biggest number)
  8. Catch (the honest counterpoint)
  9. Payoff ("which means...")
  10. CTA ("Good thing I made [freebie]. For the link, comment KEYWORD.")

Shorten beats rather than speeding up the voice. Rules:

  • Second person, contractions, short clauses joined by "and", "because", "which means". Vocabulary at the audience level in SERIES.md.
  • Name a big brand in the first sentence, even when the subject is something else.
  • Every number is concrete and sourced. One feature stack, always three items.
  • At least one real limitation, in the VO or the banners. The CTA keyword is one short word tied to the topic.
  • The last line repeats the hook's key word or number, then gives the keyword.

Write it to narration/script.json as [{"id": "l1_hook", "text": "..."}, ...], one line per sentence or clause group. Line l1_hook is SAY from the HOOK block. Mark the one or two words each line leans on (the number, the twist word, the verdict) in the staging table; papercut-voice uses them.

4. Banner track

10-18 banners, 3-7 words, ALL CAPS, one fact each, about one every 2-3 s. They never repeat the VO: read alone they are a fact sheet. Include a headline stat, a price translation, a credibility fact, availability, at least one honest caveat, at least one trust or safety fact, feature claims, at least one competitor contrast, and a range/decision banner before the CTA. Types and examples: references/storytelling.md section 3.

5. Staging table

One row per banner, written into STORYBOARD.md under the HOOK block:

t (s) | VO words under it | BANNER | SET (room + bg colour) | HERO PROP (a visual pun from the VO's concrete noun) | PRODUCT UI moment | MASCOT action and costume prop | GAG

  • A new set for every banner, background colours rotating. The set illustrates a word in the VO line, not the banner.
  • Numbers become physical props: price signs, scales, multiplier walls, gauges, meters, races. Competitors appear as pinned cards, never as the hero.
  • The mascot is the viewer's proxy and reacts. Product UI is a rough rebuild on a paper monitor or laptop with fake interactions.
  • One callback to an earlier set before the CTA. The CTA set shows a comment box typing the keyword.
  • papercut-storyboard adds the camera and transition columns.

6. Humour pass (after the script, before staging is final)

Add the GAG column. The voiceover stays straight; only the picture is silly.

  1. One format for the whole reel.
  2. One gag per banner that acts out the point, from references/comedy.md.
  3. A costume prop per idea. The mascot's own item stays on and gets one prop on top.
  4. One struggling-mascot beat, on the catch line.
  5. A prop on frame 0 that returns changed in the CTA set (I'M OUT becomes FREE).
  6. One absurd, unexplained background moment.
  7. At most one meme, rebuilt in the series' own style, never a real clip, still or audio.

No gag may cover the banner, the captions or the key number, and each must read in under 1 s.

Gate

  • 75-95 words, 22-30 s, one honest limitation, a stack of exactly three.
  • Every number and banner traces to facts.md; banners alone read as a fact sheet.
  • Last line repeats the hook's key word or number.
  • Humour pass done: costume per idea, a struggling beat, a returning prop, one absurd moment, at most one meme.

Hand-off

Next is papercut-storyboard. It reads narration/script.json and the HOOK block and staging table in episodes/NN-slug/papercut/STORYBOARD.md. For a script-only run, stop here and give the user those files. The copy-paste storytelling prompt for another tool is references/storytelling.md section 7.

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 Script AI skill do?

Turn a link into a papercut reel script. Researches a sourced facts list, writes the 75-95 word script on the 10-beat spine, a banner fact track, a staging table with one set per banner, and the humour pass. Use for "write the script", "research this repo for a reel".

Why use Papercut Script on TypingMind?

Because you install it once and use it with any model. Papercut Script 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 Script in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/0xS4N4TG/papercut-explainer/tree/main/skills/papercut-script. 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 Script?

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 Script?

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

Is the Papercut Script 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 👇