Guide · Agent skill
How to use the Data Storytelling agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The Data Storytelling skill is a set of written guidance that tells an AI agent how to turn analytics into a narrative. It is aimed at analysts, operators and anyone who has to present numbers to executives, boards or non-technical colleagues. After this guide you will know what the skill changes, how to install it, how to ask for it in plain language, and where it should not be relied on.
Data Storytellingwshobson/agents · MIT Download the skill Quick start
- 1Download the ZIP. Get the data-storytelling skill ZIP from the directory page.
- 2Choose an install location. Pick ~/.claude/skills/ for personal use or .claude/skills/ inside a project for project-only use.
- 3Unzip the skill. Extract the ZIP into the chosen folder so the skill files, including SKILL.md, are in place.
- 4Or upload in the Claude apps. In the Claude apps, open Settings, go to Capabilities, then Skills, and upload the skill.
- 5Make a natural request. Describe the audience and goal, such as an executive presentation or a data report.
- 6Check the numbers. Verify every figure against your source data before the story goes to stakeholders.
What an agent skill is and what this one changes
An agent skill is a folder containing a SKILL.md file with instructions that an AI agent can load when a task matches the skill's description. This one is called data-storytelling, and its description says it applies when you are presenting analytics to stakeholders, creating data reports or building executive presentations.
Once active, the agent structures findings as a story. It leads with the key insight, sets up context and a baseline, builds through supporting data points, and closes with recommendations and next steps. It also applies guidance on choosing clear visuals and avoiding data dumps and jargon. The skill contains no scripts or tools, so it changes how the agent frames and presents information, not what it can calculate or draw.
- Hook: a surprising insight to grab attention
- Context: the baseline the numbers are compared against
- Rising action: data points that build the case
- Climax: the key insight
- Resolution: recommendations
- Call to action: next steps
Install the Data Storytelling skill
Download the ZIP for the skill from the directory page. Unzip it into one of two locations. For a personal install available across your projects, use ~/.claude/skills/. For a skill that only applies to one project, use the .claude/skills/ folder inside that project.
If you use the Claude apps instead, open Settings, go to Capabilities, then Skills, and upload the skill there.
- Personal install: unzip into ~/.claude/skills/
- Project install: unzip into .claude/skills/ inside the project folder
- Claude apps: Settings → Capabilities → Skills, then upload
Ask for it with natural requests
You do not need a special command. Describe the job in ordinary words that match the skill's purpose, such as presenting analytics to stakeholders, writing a data report or building an executive deck. The skill's description is written to match those situations, so mention the audience and the goal in your request.
The more context you give, such as the audience, the decision they need to make and the source data, the more useful the story will be.
- Turn these Q3 sales figures into a story for the board, leading with the main finding and ending with recommendations.
- Draft a quarterly business review narrative for our operations team using these metrics.
- Rewrite this analytics report so a non-technical manager can follow it without jargon.
- Build the outline of an executive presentation on why churn rose this quarter.
Examples of what the output looks like
The reference file in the skill, references/details.md, includes a worked template for a problem-solution story. It runs from a hook, to context, to the problem, to the insight, to the solution, to expected impact, and ends with a specific call to action. This is the shape to expect from the agent: a short arc rather than a list of charts.
The example figures in that template are illustrations of the structure, not findings. Do not reuse them as real results.
- Hook: one line stating the stakes
- Context: current rate and comparison points
- Problem: the pattern the data shows
- Insight: the single finding, shown with a visual
- Solution and expected impact
- Call to action with a clear ask
Safety notes: scripts, network and checking the numbers
The skill runs no scripts and uses no network access, and it lists no required dependencies. Its files are SKILL.md and references/details.md.
The skill only shapes narrative. It does not calculate or verify statistics. A persuasive frame can make weak results look stronger than they are, so check every figure against your source data before presenting. The skill is also not a tool for producing charts or software, so build the visuals in your usual tools.
- No scripts are run by this skill
- No network use
- No listed requirements
- Verify all numbers and comparisons yourself
- Do not present a strong story built on weak results
Tips for better results
Give the agent the raw facts and the audience first, then ask for the story. Ask it to lead with the so-what and to cut anything that does not support the key insight. The skill's own guidance is to curate ruthlessly, front-load findings and explain the context before the method.
Review the output for jargon and for any claim the data does not back up.
- Name the audience and the decision they must make
- Paste the source numbers so they can be checked
- Ask for the methodology to come after the context, not before
- Ask for a plain-language version for non-technical readers
What people use it for
- Shaping a quarterly business review so the main finding comes first
- Turning a dense analytics report into a short narrative for executives
- Preparing a board presentation with context, evidence and a clear ask
- Rewriting technical metrics for non-technical team members
Frequently asked questions
Is the Data Storytelling skill free?
The source material lists the skill under the MIT licence. Check the directory page for any other terms of use before you rely on it commercially.
What do I need to run the Data Storytelling skill?
The skill lists no requirements. It runs no scripts and uses no network, so it works as written guidance inside an agent that supports skills.
Where do I put the skill folder?
Unzip it into ~/.claude/skills/ for personal use or into .claude/skills/ inside a project for project-only use. In the Claude apps, you can upload it under Settings, Capabilities, then Skills.
Will this skill create charts for me?
No. The visible files contain no scripts or tooling for chart generation. The skill guides which visuals to choose and how to avoid clutter, but you need your own charting tools to build them.
Can I trust the numbers the agent writes?
Not without checking. The skill only shapes the narrative and does not calculate or verify statistics. Compare every figure with your source data, because a persuasive frame can overstate weak results.
How is this different from a general prompt asking for a presentation?
The skill gives the agent a consistent structure: insight first, then context, evidence, recommendations and next steps, plus rules against data dumps and jargon. A general prompt may not apply those patterns unless you ask for them each time.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you are an analyst preparing executive or board presentations, a team writing quarterly business reviews, or anyone translating metrics for non-technical audiences. Pick something else if you need exact calculations, verified statistics or hands-on chart generation, since this skill only shapes how the story is told.
Download the skill