Presentation Builder logo

Presentation Builder

Community
neuromechanist
presentation-builder

This skill should be used when the user asks to "create a presentation", "make slides", "build a slide deck", "create a talk", "make a keynote", "create a Reveal.js presentation", "generate presentation slides", "make a conference talk", "create a lecture", "build a poster presentation", "create presentation JSON", or mentions presentations, slides, slide decks, Reveal.js, talk preparation, conference presentations, or lecture slides.

Overview

Publisherneuromechanist
Repositoryresearch-skills
Skill namepresentation-builder
Stars
51
Forks
9
Bundled files
4
LicenseBSD-3-Clause
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.

  • 4 bundled files

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

  • Open source

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

Installation

Install the Presentation Builder 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/neuromechanist/research-skills/tree/main/plugins/presentation/skills/presentation-builder
  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/neuromechanist/research-skills.git /tmp/research-skills
mkdir -p .claude/skills
cp -r /tmp/research-skills/plugins/presentation/skills/presentation-builder .claude/skills/presentation-builder
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Presentation Builder 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 Presentation Builder 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 Presentation Builder 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.

Presentation Builder

Create interactive Reveal.js presentations from JSON using the Agentic Presentation Builder. The builder transforms structured JSON definitions into professional, interactive web-based presentations with Mermaid diagrams, LaTeX math, syntax-highlighted code, and animated progressive reveals.

Pipeline Overview

1. Plan structure  -->  2. Author JSON  -->  3. Validate  -->  4. Serve & present
   (outline, theme)     (schema-driven)      (CLI validator)    (Vite dev server)

Prerequisites: get the builder CLI

The engine ships an apb command (subcommands validate, present, export, shoot). Two ways to run it; pick per situation. Pin the tag (#v0.1.8) for reproducibility.

Zero-setup (default, no clone). Run straight from the repo with bunx (or npx):

bash
bunx github:neuromechanist/agentic-presentation-builder#v0.1.8 validate deck.json --json

Iterative authoring / offline (recommended when validating repeatedly). Use a managed cache clone so each call does not re-resolve the git package. Resolve a builder home, cloning once if needed, then run the bun run scripts from it:

bash
APB_HOME="${APB_HOME:-$HOME/.cache/agentic-presentation-builder}"
if [ ! -d "$APB_HOME/.git" ]; then
  git clone --branch v0.1.8 https://github.com/neuromechanist/agentic-presentation-builder.git "$APB_HOME"
  (cd "$APB_HOME" && bun install)
fi
# then, e.g.:
(cd "$APB_HOME" && bun run validate -- "$(pwd)/deck.json" --json)

In the steps below, "apb <command>" means either the bunx form or bun run <command> -- from $APB_HOME. Both share one code path, so flags are identical.

Step 1: Plan the Presentation

Before writing JSON, determine:

  • Topic and audience: Shapes content depth and vocabulary
  • Slide count: 8-12 for a short talk, 15-25 for a full session
  • Theme: academic for research talks, default for general, dark for tech demos
  • Key visuals: Which slides need Mermaid diagrams, images, code blocks, or tables
  • Speaker notes: Include delivery guidance for each slide

Step 2: Author the Presentation JSON

Write a presentation.json following the schema. See references/schema-reference.md for the complete field reference and references/authoring-guide.md for best practices.

Minimal structure

json
{
  "presentation": {
    "metadata": {
      "title": "My Presentation",
      "author": "Author Name",
      "theme": "academic",
      "aspectRatio": "16:9",
      "controls": {
        "slideNumbers": true,
        "progress": true
      }
    },
    "slides": [
      {
        "id": "title",
        "layout": "title",
        "elements": [
          {
            "type": "text",
            "content": "# Presentation Title",
            "style": { "fontSize": "xxl", "alignment": "center" },
            "position": { "area": "center" }
          },
          {
            "type": "text",
            "content": "Author Name -- Conference 2026",
            "style": { "fontSize": "large", "alignment": "center", "color": "#64748B" },
            "position": { "area": "center", "order": 1 }
          }
        ]
      }
    ]
  }
}

Slide patterns

Content slide (single-column):

json
{
  "id": "key-findings",
  "layout": "single-column",
  "speakerNotes": "Emphasize the 3x improvement in latency.",
  "elements": [
    {
      "type": "text",
      "content": "## Key Findings",
      "style": { "fontSize": "xl" },
      "position": { "area": "header" }
    },
    {
      "type": "bullets",
      "items": [
        "**3x reduction** in processing latency",
        "95% accuracy on held-out test set",
        "Compatible with existing EEG pipelines"
      ],
      "position": { "area": "content" }
    }
  ]
}

Comparison slide (two-column):

json
{
  "id": "comparison",
  "layout": "two-column",
  "elements": [
    {
      "type": "text",
      "content": "## Before vs After",
      "style": { "fontSize": "xl" },
      "position": { "area": "header" }
    },
    {
      "type": "bullets",
      "items": ["Manual annotation", "Hours per session", "Inconsistent labels"],
      "position": { "area": "left" }
    },
    {
      "type": "bullets",
      "items": ["Automated pipeline", "Minutes per session", "Standardized HED tags"],
      "position": { "area": "right" }
    }
  ]
}

Diagram slide (Mermaid):

json
{
  "id": "architecture",
  "layout": "single-column",
  "elements": [
    {
      "type": "text",
      "content": "## System Architecture",
      "style": { "fontSize": "xl" },
      "position": { "area": "header" }
    },
    {
      "type": "mermaid",
      "diagram": "graph LR\n  A[Raw EEG] --> B[Preprocessing]\n  B --> C[Feature Extraction]\n  C --> D[Classification]\n  D --> E[BCI Output]",
      "position": { "area": "content" }
    }
  ]
}

Element types summary

TypeRequired fieldsKey options
texttype, contentMarkdown, LaTeX math, style, animation
bulletstype, itemsbulletStyle, nested items, animation
imagetype, srcalt, width, height, caption
mermaidtype, diagramtheme (default/dark/forest/neutral)
callouttype, contentcalloutType (tip/warning/important/note/info), title
codetype, codelanguage, caption, lineNumbers
tabletype, headers, rowscaption

Step 3: Validate

Run the CLI validator before serving:

bash
# zero-setup:
bunx github:neuromechanist/agentic-presentation-builder#v0.1.8 validate presentation.json --json
# or from the cache clone:
(cd "$APB_HOME" && bun run validate -- "$(pwd)/presentation.json" --json)

The --json flag returns structured output with valid, summary.errorCount, summary.warningCount, errors[], and warnings[]. Each issue includes code, severity, path, message, and suggestion.

Fix all schema errors. Address advisory warnings to improve slide quality (see references/authoring-guide.md for the full list and thresholds).

Step 4: Serve and Present

Use the present command. It serves the JSON directly and rewrites relative asset paths, so there is no need to copy anything into public/:

bash
# zero-setup:
bunx github:neuromechanist/agentic-presentation-builder#v0.1.8 present presentation.json --open
# or from the cache clone:
(cd "$APB_HOME" && bun run present -- "$(pwd)/presentation.json" --open)

This prints the browser URL (and accepts --port N). Press P to toggle between authoring and presentation modes, S for speaker notes, O for slide overview. See references/authoring-guide.md for the full keyboard shortcuts and delivery modes. Append &role=audience&mode=presentation to the printed URL to open a synced audience screen.

To produce a distributable file instead of serving, use export (PDF default; PPTX available):

bash
# zero-setup:
bunx github:neuromechanist/agentic-presentation-builder#v0.1.8 export presentation.json --format pdf
# or from the cache clone:
(cd "$APB_HOME" && bun run export -- "$(pwd)/presentation.json" --format pdf)

Step 5: QC every slide (don't skip)

validate passes decks that still render a blank mermaid or clipped code -- the validator cannot see rendered output. Before shipping a deck (especially a high-stakes one), screenshot every slide at full HD with shoot and look at each one:

bash
# zero-setup:
bunx github:neuromechanist/agentic-presentation-builder#v0.1.8 shoot presentation.json --out ./qc
# or from the cache clone:
(cd "$APB_HOME" && bun run shoot -- "$(pwd)/presentation.json" --out ./qc)

shoot serves the deck, drives headless Chrome through every slide, and writes one PNG per slide to --out (default 1920x1080; --width/--height to change). It disables transitions and fragments during capture so an unsettled slide-transform never fakes a right-edge clip and every animated element shows. Common fixes the QC pass catches -- avoid mermaid (renders blank here; use a table or an SVG image), keep code blocks to ~6 lines (height-capped), and put punchline callouts in area: "footer". See references/course-style.md for the full list.

Additional Resources

Reference files

  • references/schema-reference.md -- Complete field reference for all element types
  • references/authoring-guide.md -- Layout patterns, content density guidelines, theme selection, validation workflow
  • references/course-style.md -- Stable "house style" from the OSC Agentic Research Course decks: incremental bullet animations, code sections, two-column image layouts, callouts, and the title-slide block

External documentation

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 Presentation Builder AI skill do?

This skill should be used when the user asks to "create a presentation", "make slides", "build a slide deck", "create a talk", "make a keynote", "create a Reveal.js presentation", "generate presentation slides", "make a conference talk", "create a lecture", "build a poster presentation", "create presentation JSON", or mentions presentations, slides, slide decks, Reveal.js, talk preparation, conference presentations, or lecture slides.

Why use Presentation Builder on TypingMind?

Because you install it once and use it with any model. Presentation Builder 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 Presentation Builder in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/neuromechanist/research-skills/tree/main/plugins/presentation/skills/presentation-builder. 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 Presentation Builder?

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 Presentation Builder?

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

Is the Presentation Builder AI skill free?

Yes. It is published on GitHub by neuromechanist under the BSD-3-Clause 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 👇