Guide · Agent skill
How to use the Build a Dashboard agent skill (build-dashboard)
Updated · written with AI from the project's own documentation and checked against our catalogue data
The Build a Dashboard skill is a set of instructions that teaches an AI agent how to build a monitoring or exploration dashboard from your data, rather than a one-off report. It is for analysts who need a shared, repeatable KPI view and developers who want a small internal monitoring page. After following this guide you will know how to install the skill, ask for a dashboard in plain language, and check what the agent produced.
Build a DashboardXiaomiMiMo/MiMo-Code · MIT Download the skill Quick start
- 1Download the ZIP. Download the build-dashboard ZIP from this page and unzip it.
- 2Place the skill folder. Move the unzipped folder into ~/.claude/skills/ for personal use, or into a project's .claude/skills/ folder.
- 3Or upload it in the Claude apps. In the Claude apps, open Settings, go to Capabilities, then Skills, and upload the skill.
- 4Describe the dashboard. Ask for the dashboard in plain language, naming the audience, the decisions and the metrics.
- 5Choose a surface. State whether you want a BI platform, a Streamlit app or a plain HTML page, or let the agent pick the simplest supported option.
- 6Supply the data. Provide source files or query access so the agent can validate joins, duplicates and gaps.
- 7Check the QA output. Review the reconciled totals, metric definitions, freshness notes and known limitations before sharing.
What the build-dashboard skill makes the agent do
When the skill is active, the agent follows a structured process. It first identifies the audience, the decisions the dashboard supports, and the questions it must answer. It then defines each metric with its formula, time window, filters, owner, source and freshness expectation.
Before presenting any numbers as trustworthy, the agent validates joins, completeness, duplicates, missing values and refresh status. It lays out the page from overview and primary outcomes, through drivers and diagnostic cuts, to guardrails and detail tables. The agent is also asked to reconcile displayed totals against reviewed source results and to test filters, date boundaries and empty or error states.
- Surface options: an existing BI or product analytics platform, a Streamlit app, or a self-contained plain HTML/CSS/vanilla JavaScript page.
- The skill explicitly does not require React, Vite, Recharts, MCP Apps or widget runtimes.
- Delivery includes the dashboard, metric definitions, source or query files, refresh instructions and known limitations.
Install the build-dashboard skill
Download the build-dashboard ZIP from this page and unzip it. For a personal install that works across your projects, place the unzipped skill folder in ~/.claude/skills/. To scope it to one project, place it in that project's .claude/skills/ folder instead.
If you use the Claude apps, you can upload the skill instead. Open Settings, go to Capabilities, then Skills, and upload the skill there.
- Personal install: ~/.claude/skills/
- Project install: .claude/skills/ inside your project
- Claude apps: Settings → Capabilities → Skills
Trigger the skill with a natural request
You do not need a special command. Describe the dashboard you want in plain language, and say what it is for. The more concrete you are about the audience and decisions, the better the agent's metric definitions will be.
If you already have metric definitions, paste them in or point the agent to them. The skill asks the agent to read the relevant analysis and KPI workflows first when definitions are not settled.
- "Build a monitoring dashboard for weekly signups, activation and churn for the growth team."
- "Make a Streamlit dashboard of support ticket volume and response time from this CSV."
- "Create a single self-contained HTML dashboard of order revenue by region with date filters."
What you need before you start
The skill itself runs no scripts and uses no network access. What you need depends on the surface you choose. You must supply the source data, either as files or query access.
For a Streamlit app you need Python and Streamlit. For a BI platform surface you need an existing BI or product analytics platform and access to it. A plain HTML dashboard needs only a browser to view it.
- Source data files or query access supplied by you (always required).
- Python and Streamlit, only if you choose the Streamlit surface.
- An existing BI or product analytics platform with access, only if you choose that surface.
- No React or Vite is needed.
Safety notes: scripts, network and sensitive data
The skill does not run scripts and does not use the network. However, the agent you use may have its own tools, so review what it does in your session.
The skill asks the agent to avoid exposing secrets, credentials, unnecessary personal data and temporary local paths in the artifact. This is a request to the agent, not a guarantee. A person must still check what a published dashboard reveals before sharing it.
- Review any dashboard that shows personal or sensitive data before publishing it.
- Confirm that no credentials or local file paths appear in the output.
- Use the BI platform surface when you need platform-level permissions.
Tips for better dashboards
State your preferred surface and framework up front. If you do not, the agent chooses the simplest supported option, and it will not pick React because the skill avoids requiring it.
Ask for the definitions and the refresh context explicitly. Check the default date range and filter state on first load, since the skill asks that these be visible on the page.
- Name the audience and the decision the dashboard supports.
- Ask for a table alongside each chart so exact values are readable.
- Ask the agent to test empty, partial and error states.
- Use a one-off analysis instead for questions you will ask only once.
What people use it for
- A weekly KPI view for a growth team showing signups, activation and churn with definitions beside each chart.
- A small internal Streamlit page tracking support ticket volume and response time from CSV exports.
- A self-contained HTML page of order revenue by region with date range filters that opens in a browser.
- A governed dashboard built on an existing BI platform, where the agent prepares the metric definitions and query files.
Frequently asked questions
How do I install the build-dashboard skill?
Download the ZIP from this page and unzip it. Place the folder in ~/.claude/skills/ for personal use or in a project's .claude/skills/ folder. In the Claude apps, you can instead upload it under Settings, Capabilities, then Skills.
Is the build-dashboard skill free to use?
The item is published under the MIT licence. This page does not state pricing for the skill, so check the licence file in the source repository for the terms that apply.
What do I need to run it?
You need source data files or query access. Python and Streamlit are needed only for the Streamlit surface, and a BI or product analytics platform with access is needed only for that surface. The skill runs no scripts and uses no network.
Does it require React?
No. The skill explicitly avoids requiring React, Vite or Recharts. If your project must use React, state that preference, because the agent may otherwise choose plain HTML or Streamlit.
When should I use a different tool or a simple report instead?
Use a simple analysis or report for one-time ad hoc questions, since a full dashboard adds overhead. If you need a fully managed, permissioned dashboard, use your existing BI platform directly and use this skill only to prepare its definitions and queries.
Can the agent publish sensitive data by mistake?
The skill asks the agent to avoid exposing secrets, credentials and personal data, but this is not a guarantee. A human must review what the dashboard reveals before it is shared.
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
Choose this skill if you need a shared, repeatable KPI or monitoring view and want metric definitions, freshness and QA handled as part of the build. Skip it for one-off questions, or if your project requires a specific framework such as React and you are not willing to state that preference to the agent.
Download the skill