Polished Presentations logo

Polished Presentations

Community
simonmesmith
polished-presentations

Create polished executive or client presentation decks in Codex through a staged workflow: optional research, user direction, narrative outline, subagent critique, wireframe, mood boards, and final full-slide image design. Use when the user asks to make a polished presentation, turn research, notes, or dictated direction into a deck, create an executive/client deck, build a deck with mood boards, or design slides using image generation. Requires the imagegen skill for mood boards and final slide images.

Overview

Publishersimonmesmith
Repositoryagent-skills
Skill namepolished-presentations
Stars
51
Forks
5
Bundled files
2
LicenseGPL-3.0
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 simonmesmith on GitHub. Read the source before you install it.

Installation

Install the Polished Presentations 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/simonmesmith/agent-skills/tree/main/skills/polished-presentations
  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/simonmesmith/agent-skills.git /tmp/agent-skills
mkdir -p .claude/skills
cp -r /tmp/agent-skills/skills/polished-presentations .claude/skills/polished-presentations
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Polished Presentations 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 Polished Presentations 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 Polished Presentations 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.

Polished Presentations

Dependency

Load and follow the imagegen skill before generating mood boards or final slide designs.

Use the built-in image generation path by default unless the user explicitly asks for the imagegen CLI fallback.

Core Principle

Separate content creation from design production.

Do not create content and final design at the same time. First resolve the research, direction, narrative, critique, and wireframe. Only then generate mood boards and final slide images.

Workspace Structure

If the workspace already has a clear versioning convention, follow it. Otherwise create versioned folders:

text
sources/
research/
v1/
vN/

Use the next available version folder. Do not overwrite an existing version unless the user explicitly asks.

  • sources/: original or user-provided files such as RFPs, briefs, brand guidelines, PDFs, docs, spreadsheets, datasets, exported threads, and reusable brand assets.
  • research/: reusable Codex-created source notes, source indexes, extracted quotes, Slack findings, web findings, and file summaries.
  • vN/: version-specific synthesis and deck artifacts.

Each version should usually contain:

text
vN/
  source_manifest.md
  research_brief.md
  direction.md
  outline.md
  critique.md
  wireframe.pptx
  mood_boards/
  designed_slides/
  final_deck.pptx

Create vN/source_manifest.md listing the exact files, research notes, links, and datasets used for that version. Keep reusable source files in sources/; do not duplicate large source files into every version unless the source itself must be snapshotted.

Always save the content outline as a standalone Markdown file at vN/outline.md before critique and wireframing.

Workflow

1. Research

Research is optional. If the user says no research is needed, proceed from their direction, notes, or dictated thoughts.

When research is needed, gather relevant source material from the user-provided context, local files, connected tools, Slack channels, web sources, docs, data, or notes.

Save reusable findings in research/. Save a version-specific synthesis in vN/research_brief.md.

Track:

  • facts and examples that should inform the deck
  • useful quotes or source excerpts
  • open questions, weak evidence, and assumptions
  • source links, file paths, dates, and relevant context

Keep research notes separate from slide copy.

2. Direction

Capture the user's point of view in vN/direction.md.

Treat dictation or messy idea dumps as a strong input format. Preserve the user's intent, emphasis, and language, then organize it into usable direction.

Include:

  • target audience and what they care about
  • purpose of the presentation
  • desired audience reaction or decision
  • thesis or emerging point of view
  • tone, stakes, constraints, and must-say points
  • what to avoid

If direction is thin, ask concise questions before outlining unless the user has asked you to proceed with assumptions.

3. Outline

Create vN/outline.md as the working source of truth for the deck's content.

Include:

  • title or working title
  • audience and purpose
  • thesis
  • slide-by-slide outline
  • each slide's role in the story
  • audience-facing slide content
  • evidence or data needed for the slide
  • notes for presenter emphasis, when helpful
  • open questions or unresolved content risks

Keep the outline audience-facing. Do not include meta-content that explains the workflow, the prompt, the AI, the research process, or the fact that a deck is being created unless that is intended for the actual presentation audience.

4. Critique

Critique the outline before wireframing.

If the environment supports subagents, use them. Keep critique subagents clean and objective: give them the outline, audience, purpose, and necessary source context, but do not pass the full working conversation, your intended revisions, or leading conclusions. Ask each subagent for independent critique rather than validation.

Cover exactly these perspectives:

  • Narrative / story flow: clear arc, tension, payoff, logical progression, and slide-to-slide momentum. A Nancy Duarte-style presentation strategist perspective is useful here.
  • Content and conciseness: usefulness, brevity, non-repetition, presenter-friendliness, and whether each slide earns its place. A Strunk & White or William Zinsser-style clarity perspective is useful here.
  • Target audience resonance: relevance, credibility, likely objections, and whether the intended audience will care.

If subagents are unavailable, simulate these separate reviewer perspectives yourself and state that the critique was simulated locally.

Save critique notes and resulting changes in vN/critique.md. Then revise vN/outline.md before proceeding.

5. Wireframe

Create a simple 16:9 content wireframe deck at vN/wireframe.pptx.

Use minimal design only: clean layouts, black text when reasonable, simple shapes, and clear hierarchy. The wireframe is for content control, not final visual polish.

Requirements:

  • Include only audience-facing content.
  • Make each slide content-complete enough to guide image generation.
  • Use concise slide copy intended for presentation, not reading.
  • Include clear chart titles, legends, axis labels, and data labels on every chart.
  • Add labels to key values and data points that must survive final image generation.
  • Avoid ambiguous charts, unlabeled axes, and decorative design decisions that should be handled later.

6. Mood Boards

Load the imagegen skill.

Generate three meaningfully different mood board directions in vN/mood_boards/. Each direction should be a distinct territory relevant to the subject matter, not three minor variations of the same style.

Each direction should explore:

  • style and visual language
  • color palette
  • typography feel
  • composition and layout rhythm
  • imagery, texture, and atmosphere
  • emotional tone

Use brand guidelines, brand colors, required fonts, palette constraints, or user-provided references when available. Ask the user to pick, reject, or combine directions before final slide design unless they have already given clear design direction.

7. Design

Load and follow the imagegen skill.

For each wireframe slide, generate one full-bleed 16:9 image in the chosen mood board style. Save final selected images in vN/designed_slides/.

Default design rule:

Generate the entire slide as a single image, including background, text, charts, labels, layout, and visual design. Do not generate only a background. Do not overlay live PowerPoint text or charts unless the user explicitly chooses a hybrid editable workflow.

The image must be a true 16:9 slide image, full bleed edge-to-edge, with no white border, margin, frame, mat, canvas edge, screenshot frame, or padding around the slide.

Each slide image prompt should include:

  • production slide number for tracking and the visible slide title
  • exact audience-facing text to render
  • chart data and required labels, if any
  • the chosen mood board/style direction
  • full-bleed 16:9 slide composition
  • cohesion with the rest of the deck
  • variation from adjacent slides so the deck is not visually monotonous

If the user does not want visible slide numbers, do not render slide numbers on the slide image. Still track slide numbers in filenames and production notes.

Compile the designed images into vN/final_deck.pptx as image-only full-slide slides.

Failure Modes To Prevent

  • Do not write content and design at the same time.
  • Do not generate final slide backgrounds and then add live text by default.
  • Do not mix generated imagery with live text or charts unless the user explicitly asks for a hybrid editable deck.
  • Do not include fourth-wall content such as "this slide explains," "the deck will show," "based on the research," "AI-generated," or notes about the workflow unless the presentation audience is supposed to see that.
  • Do not omit data labels from wireframe charts.
  • Do not leave charts, axes, legends, or key values ambiguous before image generation.
  • Do not allow white borders, margins, frames, mats, or padding around generated slide images; every final slide image must be true 16:9 full bleed.
  • Do not make every final slide composition look the same; enforce cohesion with variation.
  • Do not overwrite prior versions or source artifacts.
  • Do not leave project-bound generated images only under the default imagegen output location; move or copy final assets into the active version folder.

Tradeoffs To Note

Image-only final decks can look much more polished, but they are harder to edit, less suitable for object-level animation, and can produce large files.

If the user needs a highly editable final deck, offer a hybrid workflow as an explicit alternative: generated backgrounds with editable foreground text/charts. This workflow has not been optimized here, so use it only when editability is more important than image-native polish.

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 Polished Presentations AI skill do?

Create polished executive or client presentation decks in Codex through a staged workflow: optional research, user direction, narrative outline, subagent critique, wireframe, mood boards, and final full-slide image design. Use when the user asks to make a polished presentation, turn research, notes, or dictated direction into a deck, create an executive/client deck, build a deck with mood boards, or design slides using image generation. Requires the imagegen skill for mood boards and final slide images.

Why use Polished Presentations on TypingMind?

Because you install it once and use it with any model. Polished Presentations 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 Polished Presentations in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/simonmesmith/agent-skills/tree/main/skills/polished-presentations. 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 Polished Presentations?

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 Polished Presentations?

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

Is the Polished Presentations AI skill free?

Yes. It is published on GitHub by simonmesmith under the GPL-3.0 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 👇