Not in the top 1,000 by GitHub stars
andreworia/claude-powerpoint-skills
12 Claude skills for building board-ready PowerPoint decks: storyline, action titles, chart selection and partner-grade review. Built by Oria — AI for complex professional slides.
History
Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.
View weekly changes
skills/action-title-writer/SKILL.md
Week of Nov 2, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 45f80c2fab1e
No earlier version was observed for comparison.
skills/board-narrative-scqa/SKILL.md
Week of Nov 2, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 45f80c2fab1e
No earlier version was observed for comparison.
skills/chart-to-message-mapper/SKILL.md
Week of Nov 2, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 45f80c2fab1e
No earlier version was observed for comparison.
skills/chart-type-selection/SKILL.md
Week of Nov 2, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 45f80c2fab1e
No earlier version was observed for comparison.
skills/color-encoding-for-data/SKILL.md
Week of Nov 2, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 45f80c2fab1e
No earlier version was observed for comparison.
skills/data-ink-reducer/SKILL.md
Week of Nov 2, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 45f80c2fab1e
No earlier version was observed for comparison.
skills/decision-and-ask-page/SKILL.md
Week of Nov 2, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 45f80c2fab1e
No earlier version was observed for comparison.
skills/executive-summary-page/SKILL.md
Week of Nov 2, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 45f80c2fab1e
No earlier version was observed for comparison.
skills/framework-slide-designer/SKILL.md
Week of Nov 2, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 45f80c2fab1e
No earlier version was observed for comparison.
skills/ghost-deck-skeleton/SKILL.md
Week of Nov 2, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 45f80c2fab1e
No earlier version was observed for comparison.
skills/partner-review-checklist/SKILL.md
Week of Nov 2, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 45f80c2fab1e
No earlier version was observed for comparison.
skills/waterfall-and-bridge-charts/SKILL.md
Week of Nov 2, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 45f80c2fab1e
No earlier version was observed for comparison.
Heat
1.7- Star growth, last 7 daysNot available
- Star growth, last 24 hoursNot available
- Rank movement, last 7 daysNot available
- Commits, last 30 daysNot available
- Accounts asking about it through MCP, last 7 daysNot available
- Hot votes this weekNot available
Heat never changes the order by stars, and sponsorship never touches it. How heat works
Stars over time
30-day change not yet available: not enough history.
Individual skills in this repository
- action-title-writer
skills/action-title-writer - board-narrative-scqa
skills/board-narrative-scqa - chart-to-message-mapper
skills/chart-to-message-mapper - chart-type-selection
skills/chart-type-selection - color-encoding-for-data
skills/color-encoding-for-data - data-ink-reducer
skills/data-ink-reducer - decision-and-ask-page
skills/decision-and-ask-page - executive-summary-page
skills/executive-summary-page - framework-slide-designer
skills/framework-slide-designer - ghost-deck-skeleton
skills/ghost-deck-skeleton - partner-review-checklist
skills/partner-review-checklist - waterfall-and-bridge-charts
skills/waterfall-and-bridge-charts
How it works
action-title-writer
When your agent uses it
Rewrite topic-label slide titles into headline action titles that state the so-what. Use when titles name the subject instead of delivering the message.
What it loads into context
Frontmatter
37 tokensAlways available to the agent
SKILL.md
890 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- What it does
- Method
- Inputs
- Output format
- Example
Files
SKILL.mdcore4,084 bytes
Read SKILL.md
# Action Title Writer
## When to use
Use this when slide titles read like file names: "Q3 Revenue", "Market Overview", "Competitive Landscape". They tell the reader what the slide is about but not what it says. You need every title rewritten as an action title, a full sentence that states the slide's single message, so a reader who reads only the titles still gets the whole argument.
## What it does
It converts each topic-label title into an action title: a complete declarative sentence carrying the slide's so-what. Read top to bottom, the action titles alone form the storyline of the deck. The body of the slide becomes proof of the title rather than a place to hunt for the point.
## Method
An action title states the conclusion of the slide, not its contents. The test is the "title-only read": if someone reads only the titles of every page, they should reconstruct the argument. Topic labels fail this test; action titles pass it.
1. Find the one message. For each slide, ask what single thing this page proves. If it proves two things, it is two slides, or one of the two is really backup. The action title carries that one message and nothing else.
2. Write it as a full sentence with a verb. "Revenue" becomes "Revenue grew 12 percent, but all of it came from one account". A label has no verb and no claim. An action title has both. The verb is where the so-what lives, so choose a real one (grew, fell, shifted, concentrated, lagged), not a placeholder (relates to, concerns).View the complete source on GitHubboard-narrative-scqa
When your agent uses it
Structure the through-line of a board pack using SCQA and the Pyramid Principle so the deck argues one clear story instead of reporting disconnected pages.
What it loads into context
Frontmatter
39 tokensAlways available to the agent
SKILL.md
1,222 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- What it does
- Method
- Inputs
- Output format
- Example
Files
SKILL.mdcore5,687 bytes
Read SKILL.md
# Board Narrative (SCQA)
## When to use
Use this skill before you build any individual board page, when you have a pile of updates, numbers, and asks but no spine connecting them. It is also the right skill when a draft pack reads like a stack of department reports rather than a single argument the board can follow and act on.
## What it does
It produces the narrative skeleton for the whole board pack: a one-line governing thought, an SCQA frame for the opening, and a Pyramid Principle tree that orders every downstream page under that governing thought. The output tells you what each page must say in one sentence and in what order, so the pack reads as one argument from cover to appendix.
## Method
The skill runs the Minto Pyramid Principle on top of an SCQA opening. Work the steps in order.
1. Fix the audience and the decision. A board pack exists to inform oversight and to extract specific decisions. Write down who is in the room (full board, a committee, the chair) and the one or two decisions you need from them. If there is no decision, the pack is a read-ahead, and you say so.
2. Build the SCQA opening.View the complete source on GitHubchart-to-message-mapper
When your agent uses it
Turns a raw chart into a message-driven exhibit by adding the action title, the one highlight, the labels that matter, and removing everything that does not serve the point.
What it loads into context
Frontmatter
44 tokensAlways available to the agent
SKILL.md
705 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- What it does
- Method
- Inputs
- Output format
- Example
Files
SKILL.mdcore3,243 bytes
Read SKILL.md
# Chart-to-Message Mapper
## When to use
Use this when you have a correct chart that does not communicate, for example a default Excel chart pasted onto a slide. Use it when a reader has to study a chart to find the point. Use it after exhibit selection, to finish the exhibit so the message is unmissable.
## What it does
It transforms a neutral chart into a persuasive exhibit. It attaches the message as a title, directs the eye to the single insight, labels only what matters, and strips the chart junk so the data carries the argument by itself.
## Method
The principle is that an exhibit should make its point in under five seconds, with no narration needed.
1. Write the takeaway as the chart title. The chart gets the action title, not "Revenue by region". The title states the conclusion the chart proves.
2. Pick the one thing to see. Decide the single element the reader must notice, one bar, one inflection, one gap. Everything else exists only as context.View the complete source on GitHubchart-type-selection
When your agent uses it
Picks the right chart type for your data and your message using a decision tree, so the chart proves the point instead of just showing numbers.
What it loads into context
Frontmatter
36 tokensAlways available to the agent
SKILL.md
1,054 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- What it does
- Method
- Inputs
- Output format
- Example
Files
SKILL.mdcore4,841 bytes
Read SKILL.md
# Chart Type Selection
## When to use
Use this skill first, before you build anything. You have a table of numbers and a point to make, but you are not sure whether it should be a bar, a line, a scatter, a waterfall, or something else. Run this skill to convert "what data do I have" plus "what am I trying to say" into one defensible chart type.
## What it does
It returns a single recommended chart type, a short reason tied to your message, and two runner-up options with the trade-offs. It also flags when a chart is the wrong tool and a table or a single number would land better. The output is structured so you can drop the recommendation straight into the build step.
## Method
The skill runs a message-first decision tree. The driving question is never "what does the data look like" alone. It is "what is the one sentence the audience should remember, and which encoding proves that sentence fastest."
Step 1. State the message in one sentence.
Force the message into a single declarative claim, for example "European margin fell while volume held" or "two segments drive eighty percent of revenue." If you cannot write the sentence, stop. The chart cannot be chosen until the claim exists.
View the complete source on GitHubcolor-encoding-for-data
When your agent uses it
Applies color to a chart so it encodes meaning and directs attention, using preattentive contrast, sequential and categorical palettes, and accessibility checks.
What it loads into context
Frontmatter
40 tokensAlways available to the agent
SKILL.md
879 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- What it does
- Method
- Inputs
- Output format
- Example
Files
SKILL.mdcore4,147 bytes
Read SKILL.md
# Color Encoding for Data
## When to use
Use this on any chart once it is built and labeled. Color is the most misused element of a chart. Done well, it makes the message jump out before the reader reads a word. Done badly, it turns a chart into a rainbow that hides the point.
## What it does
It returns a color plan: the base palette, which element gets the accent, the palette type for the data (categorical, sequential, or diverging), and an accessibility check. The plan ties every color to a reason.
## Method
Step 1. Start from one color, add color only to mean something.
The default is a single muted color for all marks. Add a second color only to encode a real distinction: the focus category, increases versus decreases, above versus below target. Color with no meaning is decoration and should be removed.
Step 2. Use preattentive contrast to direct attention.
The eye locks onto the one element that differs in color before conscious reading begins. Exploit this: gray every mark except the one the message names, and give that one the accent color. A single accent bar in a field of gray is read faster than any label.View the complete source on GitHubdata-ink-reducer
When your agent uses it
Strips non-data ink from charts and slides using the data-ink ratio principle, so the remaining marks all carry meaning and the slide reads cleaner.
What it loads into context
Frontmatter
40 tokensAlways available to the agent
SKILL.md
676 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- What it does
- Method
- Inputs
- Output format
- Example
Files
SKILL.mdcore3,105 bytes
Read SKILL.md
# Data-Ink Reducer
## When to use
Use this when a chart or slide looks busy, heavy, or cluttered without being more informative. Use it on inherited templates loaded with gridlines, borders, fills, and shadows. Use it as the last pass on any exhibit before review.
## What it does
It removes every mark that does not carry information, raising the share of ink that actually represents data. The result is lighter, faster to read, and more credible, because nothing distracts from the numbers.
## Method
The data-ink ratio is the proportion of a graphic's ink devoted to data rather than decoration. The goal is to maximize it without erasing useful context.
1. Inventory the ink. List every visual element on the exhibit: bars, lines, gridlines, borders, fills, shadows, legends, axis titles, tick marks, background, icons. Mark each as data ink (represents a value) or non-data ink (decoration or redundant structure).
2. Erase non-data ink first. Remove drop shadows, 3D effects, gradient fills, heavy borders, and background fills. These never carry information and always add visual weight.View the complete source on GitHubdecision-and-ask-page
When your agent uses it
Build the decision page that converts the recommendation into a specific, votable ask, with options considered, the recommended option, and what approval commits the board to.
What it loads into context
Frontmatter
42 tokensAlways available to the agent
SKILL.md
981 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- What it does
- Method
- Inputs
- Output format
- Example
Files
SKILL.mdcore4,793 bytes
Read SKILL.md
# Decision and Ask Page
## When to use
Use this skill for the page that asks the board to decide. It is the right skill whenever the pack needs an approval, a mandate, a budget, or a go or no-go, and you want the board to leave the meeting with a clear, recorded decision rather than a vague sense of direction.
## What it does
It produces a decision page that states the question, lays out the options that were genuinely considered, names the recommended option with its rationale, and spells out exactly what approving it commits the board to. It turns "we think we should" into a motion the board can carry.
## Method
The skill runs a structured options-to-recommendation method. Boards approve choices, not ideas, so the page must show the choice.
1. State the decision question in one line, phrased as the board would vote on it. "Approve the reallocation of capital across the initiative portfolio" is votable. "Discuss our cost program" is not.
2. Establish the decision criteria up front. Name the two to four criteria the options are judged against (for example, margin impact, execution risk, time to effect, reversibility). Criteria stated after the recommendation look like rationalization; state them first.View the complete source on GitHubexecutive-summary-page
When your agent uses it
Build the one-page executive summary that opens a board pack, stating the governing thought, the three to five proof points, and the ask so a director who reads only this page is fully informed.
What it loads into context
Frontmatter
47 tokensAlways available to the agent
SKILL.md
908 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- What it does
- Method
- Inputs
- Output format
- Example
Files
SKILL.mdcore4,219 bytesSKILL.mdcore4,219 bytes
Read SKILL.md
# Executive Summary Page
## When to use
Use this skill once the board narrative is set and you need the single page that opens the pack. It is the right skill whenever a director must be able to read one page, understand the whole story, and know what is being asked of them, even if they read nothing else.
## What it does
It produces a self-contained executive summary page: an action title carrying the governing thought, a short framing line, three to five proof points each with a number and a so-what, and a clearly marked ask. The page mirrors the pyramid built by the board-narrative skill and works as the cover argument for the entire pack.
## Method
The skill applies the Pyramid Principle to a single page, with answer-first discipline.
1. Lead with the answer. The action title is the governing thought as a full sentence. The reader gets the conclusion in the first line, before any supporting detail.
2. Write one framing line under the title. It states the period and the lens (for example, "Q2 results against the board-approved plan"). It does not argue; it orients.View the complete source on GitHubframework-slide-designer
When your agent uses it
Designs a clean framework slide (2x2, value chain, process flow, pyramid, matrix) that communicates a structured idea without clutter or false precision.
What it loads into context
Frontmatter
40 tokensAlways available to the agent
SKILL.md
749 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- What it does
- Method
- Inputs
- Output format
- Example
Files
SKILL.mdcore3,430 bytes
Read SKILL.md
# Framework Slide Designer
## When to use
Use this when the point is a structure rather than a number: a positioning 2x2, a value chain, a maturity model, an operating model, a phased roadmap. Use it when a framework slide is overloaded with boxes and arrows. Use it when a partner wants "the classic layout" for a concept.
## What it does
It lays out a conceptual framework so the structure reads instantly: the right archetype, balanced spacing, a clear reading path, and just enough labelling. It avoids the common failure of a framework that looks busy and says little.
## Method
A framework slide communicates a relationship between ideas. Pick the archetype that matches the relationship, then design it with restraint.
1. Name the relationship. Two crossing dimensions implies a 2x2 matrix. A linear sequence implies a process flow or value chain. A hierarchy implies a pyramid or tree. A set of parallel components implies columns or a wheel. Maturity over levels implies a stairway. Choose the one that fits.
2. Define the axes or stages precisely. For a 2x2, label both axes with a clear low-to-high meaning and name all four quadrants. For a flow, name each stage with a verb and keep stages MECE. Ambiguous axes are the top reason a 2x2 fails.View the complete source on GitHubghost-deck-skeleton
When your agent uses it
Produces a ghost deck (skeleton) of empty slides with action titles and exhibit placeholders, so the storyline is agreed before any time is spent on design.
What it loads into context
Frontmatter
42 tokensAlways available to the agent
SKILL.md
636 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- What it does
- Method
- Inputs
- Output format
- Example
Files
SKILL.mdcore2,942 bytes
Read SKILL.md
# Ghost Deck Skeleton
## When to use
Use this at the start of building, after the storyline is set but before any real slide is designed. Use it to align a team or a partner on the argument cheaply, before anyone spends hours formatting. Use it when scope is unclear and you want to lock the page count and the flow first.
## What it does
It produces a ghost deck, also called a skeleton deck or dummy deck: every slide present, each with its action title written and a sketch of the intended exhibit, but no real design or final data. It makes the storyline visible and editable before expensive work begins.
## Method
A ghost deck is the standard top-tier-consulting way to de-risk a deck. Build it slide by slide from the storyline.
1. One slide per leaf. Take the storyline tree and create one ghost slide for each evidence leaf, plus section dividers and the executive summary. Keep them in presentation order.
2. Write the action title now. Each ghost slide gets its final action title up front. The title chain is the deliverable being reviewed. If the titles do not argue the case, fix the storyline, not the slides.View the complete source on GitHubpartner-review-checklist
When your agent uses it
Runs a deck through a partner-grade review checklist covering story, titles, exhibits, layout, and consistency, returning a prioritized fix list before review.
What it loads into context
Frontmatter
38 tokensAlways available to the agent
SKILL.md
721 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- What it does
- Method
- Inputs
- Output format
- Example
Files
SKILL.mdcore3,359 bytesSKILL.mdcore3,359 bytes
Read SKILL.md
# Partner Review Checklist
## When to use
Use this before sending a deck to a partner, MD, or client. Use it as a self-review on your own deck after building. Use it when you have an hour before a review and need to know what to fix first.
## What it does
It evaluates a deck against the standard a senior reviewer applies and returns a prioritized list of fixes. It checks the argument first, then the slide craft, so the most important problems surface before cosmetic ones.
## Method
A partner reviews top-down: story before slides, message before format. Apply the checklist in that order so the prioritized list reflects real impact.
1. Story and answer. Is there one clear recommendation, stated up front? Does the deck answer the audience's actual question? Read the executive summary alone; can the decision be made from it?
2. Title chain. Read the action titles in sequence. Do they argue the case with no gaps? Is every title a claim, not a topic? Does each title match what its slide proves?View the complete source on GitHubwaterfall-and-bridge-charts
When your agent uses it
Builds waterfall (bridge) charts that explain how a starting value became an ending value through a sequence of additions and subtractions.
What it loads into context
Frontmatter
36 tokensAlways available to the agent
SKILL.md
910 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- What it does
- Method
- Inputs
- Output format
- Example
Files
SKILL.mdcore4,176 bytes
Read SKILL.md
# Waterfall and Bridge Charts
## When to use
Use this when the message is "how did we get from A to B." A waterfall, also called a bridge, decomposes a change into its drivers: opening value, a sequence of positive and negative moves, and the closing value. It is the standard chart for profit bridges, variance analysis, headcount changes, and price-volume-mix.
## What it does
It returns a fully specified waterfall: the bar sequence, which bars are totals versus deltas, the floating-bar construction, color rule for increases and decreases, connector lines, and labeling. It makes the drivers of a change legible at a glance.
## Method
Step 1. Define the start, the end, and the drivers.
List the opening value, each driver that moves it (with sign and size), and the closing value. The drivers must sum from opening to closing exactly. If they do not, there is a missing bridge item; surface it as an explicit "other" or "unexplained" bar rather than hiding the gap.
Step 2. Order the drivers deliberately.
Order tells a story. Common orders are largest driver first, or a logical sequence (volume, then price, then mix, then cost). Group all increases then all decreases only if that aids the narrative. Keep the order stable if you show the same bridge across periods.View the complete source on GitHubSkills included
- action-title-writer
- board-narrative-scqa
- chart-to-message-mapper
- chart-type-selection
- color-encoding-for-data
- data-ink-reducer
- decision-and-ask-page
- executive-summary-page
- framework-slide-designer
- ghost-deck-skeleton
- partner-review-checklist
- waterfall-and-bridge-charts
Repository facts
- License
- MIT
- Created
- Sep 8, 2026
- Last pushed
- Sep 8, 2026
- Forks
- 7
- Open issues
- 0
- Project activity
- Active
Before you install
Instructions onlyAutomated signals, not a guarantee
Checked Nov 5, 2026
Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.
Show 2 notes
skills/executive-summary-page/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedskills/partner-review-checklist/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzed
- No broad tool permissions
- No dangerous shell commands
- No network access
- No hidden or obfuscated text
- No instruction overrides
- No executable scripts
- No access to secrets
- No requests to vote in Hot
How the signals fit together
Facts from a static read of the files at the pinned commit. They describe what the code does, not why.
Some files could not be read; the facts below still hold.
No combined signals at this commit.
Works with
| Agent | Status | Evidence | Install to |
|---|---|---|---|
| Claude Code | Compatible | Standard SKILL.md skills/action-title-writer/SKILL.md | .claude/skills/<skill-name>/ |
| Codex | Compatible | Standard SKILL.md skills/action-title-writer/SKILL.md | .agents/skills/<skill-name>/ |
| Cursor | Compatible | Standard SKILL.md skills/action-title-writer/SKILL.md | .cursor/skills/<skill-name>/ |
| Gemini CLI | Compatible | Standard SKILL.md skills/action-title-writer/SKILL.md | .gemini/skills/<skill-name>/ |
| GitHub Copilot | Compatible | Standard SKILL.md skills/action-title-writer/SKILL.md | .github/skills/<skill-name>/ |
| OpenCode | Compatible | Standard SKILL.md skills/action-title-writer/SKILL.md | .opencode/skills/<skill-name>/ |
- Native The repository ships a manifest, folder or frontmatter field for this agent.
- Compatible A standard SKILL.md that this agent can read.
- Unknown No evidence either way.
Derived automatically from paths, manifests and frontmatter. Not tested on each agent.