Brag Slim logo

Brag Slim

OrganizationPopular
latent-spaces
brag-slim

Turn a project directory or a website URL into a short, shareable launch video with music, motion, and share copy. One file, no bundled assets — built entirely by the model with the tools already on the machine. Use when someone says "/brag-slim", "let's /brag about this", "brag about <url>", "make a launch video", or wants to show off what they built. If the /brag skill is also installed, let /brag handle those phrases; it hands off here on Opus 5.5.

Overview

Publisherlatent-spaces
Repositorybrag
Skill namebrag-slim
Stars
12.5K
Forks
863
Bundled files
Instructions only
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.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by latent-spaces on GitHub. Read the source before you install it.

Installation

Install the Brag Slim 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/latent-spaces/brag/tree/main/skills/brag-slim
  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/latent-spaces/brag.git /tmp/brag
mkdir -p .claude/skills
cp -r /tmp/brag/skills/brag-slim .claude/skills/brag-slim
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Brag Slim 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 Brag Slim 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 Brag Slim 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.

/brag-slim

You built it. Now brag. You make the whole video yourself — story, visuals, audio, render — with whatever tools are on the machine.

Whatever the tone, it should feel like a modern, slick, polished launch video: nothing on screen or in the soundtrack that doesn't earn its place.

Usage: /brag-slim [input] [options]. Options (flags or plain language):

OptionDefault
--tone <preset or freeform>inferred; default if nothing clearly fits
--format landscape|vertical|squarelandscape (1920×1080; vertical 1080×1920, square 1080×1080), 30fps
--duration <s>about 20s

Write the deliverables to brag-output/ in the current directory (timestamped brag-output-YYYY-MM-DD-HHmmss/ if it already exists). Keep every intermediate file (frames, downloads, scripts, stems) in a work/ subfolder inside it.

1. Inspect

First decide what the input is, then gather material from it. Only the source changes; everything from the questions below onward is the same for every input.

InputHow to recognize itWhere the material comes from
ProjectNo input given, and the current directory is a projectThe code
WebsiteAn http(s):// URL, or a bare domain like example.comThe live site

If the input matches no type, or there's no input and the current directory isn't a project, ask the user what to brag about.

Project

Read the code: the main page, styles (exact colors and fonts), README, routes and key components. The best material is the product in use — find its 2–3 beats: entry → key action → result.

You have the source, so use it directly: import or render the project's real components, stylesheets, fonts, images and animations in the video instead of rebuilding them.

Website

Get the site as a visitor sees it. Many sites build their page with JavaScript, so a plain download can come back as an almost empty shell. If it does, load the page in a headless browser to get the rendered result. Dismiss cookie banners and other overlays, and scroll section by section, since content that animates in on scroll stays blank in a single full-page capture.

  • Copy: headline, tagline, section headings, feature names, calls to action, testimonials. Also check the title, meta description and social-preview tags.
  • Identity: exact colors from the site's CSS and the fonts it loads.
  • Visuals: the logo, product screenshots, hero images, demo videos. Download the ones you'll use into work/.
  • Screenshots: capture the page at the video's aspect ratio to understand the layout. In the video, reuse the site's real markup, CSS and assets and animate those, rather than panning over flat screenshots.
  • The product in use: check the demo videos, how-it-works sections and linked docs for the entry → key action → result flow.

Then, for every input

Before planning, answer: What is it (one sentence)? Who is it for, and what does it do for them? What sets it apart? What's the most impressive or funniest claim? What's the visual hook? What real UI or flow should be shown? What tone fits? What's the one-line share caption?

2. Plan

Write brag-plan.md: the angle, the hook, 2–3 highlights, the punchline, tone, visual identity, and a scene-by-scene storyboard with durations that sum to the target.

If the user points at one part — a new version, a new feature, one angle — make this the focus of the video.

Shape: Hook (2–3s) → Reveal (2–4s) → 2–3 sharp highlights → Punchline/outro (2–4s). A starting shape, not a template.

Creative laws

  • Short. 15–25 seconds; 18–22 is the sweet spot.
  • Clear to a stranger. After one viewing, someone who's never heard of it knows what it does, who it's for, and how to get it. Lead with that, not with how it's built.
  • The hook is everything. The first 2 seconds decide whether anyone keeps watching. Plan it first.
  • Show the thing. Reuse the real thing from the source — its UI, components, copy, images, videos and animations — rather than re-creating it. Rebuild only what you can't reuse. Prefer the working app doing its job over a landing page describing it. Small illustrative UI text is fine when showing the product in use (a filename, an "Exported" toast); invented claims, numbers, or testimonials are not. Never abstract filler.
  • Specific. It must feel made for this exact project. Use its own copy and claims; no generic SaaS language ("streamline your workflow" is banned).
  • Readable. Pace comes from motion and cuts, not from pulling text away early. Any line the viewer is meant to read stays fully visible and settled long enough to read it (roughly 0.3s per word), counted from when the whole line is on screen. Text that's only texture doesn't need to be read.
  • Make it alive. Things that appear one by one, simulated clicks, swipes, and typing beat static slides.
  • Funny earns its place. Humor comes from the project's own absurdity, not from trying.
  • Every frame postable. Any frozen frame should be worth sharing.

Tones

Presets are defaults; freeform direction ("fake Series A launch from 2016") refines or overrides them.

ToneFeelPacing / transitions
defaultPunchy, playful, clean4–5 scenes; soft transitions
polishedSerious, elegant, restrained3–4 scenes, long holds; soft fades
yc-parodyDeadpan startup launch, played straight4–5 scenes, one claim each; hard cuts
chaoticFAST, LOUD, ALL CAPS6–8 scenes, some under 2s; flash/zoom cuts
deadpanCalm, dry, nothing is a joke3–4 scenes, big empty space; slow fades
cinematicTrailer-scale, epic claims4–5 scenes, big type; dramatic wipes
app-storeClean feature cards4–6 scenes; smooth slides

Sound

Write the music and sound effects as one piece: effects in the same key and the same space as the music, blended in rather than laid on top. Give it a basic, proper mix, the way a real track is mixed: effects sit softly under the music, nothing harsh or spiky, and repeated little sounds stay in the background.

3. Build, check, render

Build it with whatever works on this machine. If you draw the video in a browser, make every frame a pure function of time and wait for fonts and images to load before capturing each one.

Before the full render, look at stills from every scene and from mid-transition, and fix overflow, collisions, and low contrast. A plain crossfade between two busy layouts makes a muddy double exposure; stagger it (old content out, then new content in) or dip through the background. Then render brag.mp4.

4. Deliver

  • Poster: pull the strongest settled frame (text fully in, not mid-transition) to brag.jpg, and bake it in as frame 0 of brag.mp4 so every platform's thumbnail shows it. Replace frame 0 rather than adding a frame, so the duration and audio sync stay the same.
  • share-copy.txt: 1–3 sentences, postable as-is, specific, matching the tone. No "excited to share."
  • Tell the user where the video and copy are, give one sentence on the creative angle, and offer to re-roll a scene or try another tone.

Frequently asked questions

What does the Brag Slim AI skill do?

Turn a project directory or a website URL into a short, shareable launch video with music, motion, and share copy. One file, no bundled assets — built entirely by the model with the tools already on the machine. Use when someone says "/brag-slim", "let's /brag about this", "brag about <url>", "make a launch video", or wants to show off what they built. If the /brag skill is also installed, let /brag handle those phrases; it hands off here on Opus 5.5.

Why use Brag Slim on TypingMind?

Because you install it once and use it with any model. Brag Slim 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 Brag Slim in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/latent-spaces/brag/tree/main/skills/brag-slim. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Brag Slim?

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 Brag Slim?

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

Is the Brag Slim AI skill free?

Yes. It is published on GitHub by latent-spaces 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 👇