Use this skill whenever the user asks for a chart, graph, plot, or visualization of data — 'chart this by month', 'show me a breakdown by category', 'visualize the trend'. Covers picking the right chart type for the question being asked, labeling/readability rules, and which plotting library to reach for. Pair with data-analysis for the underlying data prep.
The English summary is being prepared. Showing the original description.
See how to installReview the instructions the agent will receive before you install.
matplotlib, seaborn, plotly, and kaleido are already installed in the sandbox — no install_packages call needed for any workflow below.
Pick the chart type from the question being asked, not from habit — a bar chart is not the universal default:
| Question shape | Chart type |
|---|---|
| Comparing a metric across categories ("revenue by region") | Bar chart (horizontal if category labels are long) |
| Trend over time ("revenue by month") | Line chart |
| Distribution of a single variable ("how are order sizes distributed") | Histogram, or box plot for comparing distributions across groups |
| Relationship between two numeric variables ("does price correlate with rating") | Scatter plot |
| Composition of a whole ("market share by segment") | Stacked bar chart — prefer this over a pie chart once there are more than ~4-5 slices, since angle/area comparisons get hard to read past that; a pie chart is defensible for 2-4 slices where the "parts of a whole" framing is the entire point |
If the question doesn't clearly map to one of these, default to a bar or line chart (whichever fits the data shape) rather than reaching for something more exotic — a chart the user immediately understands beats a more "interesting" one they have to puzzle over.
"Revenue ($K)", not just "Revenue") — a chart with unlabeled axes forces the viewer to guess what they're looking at."viridis"/"cividis" colormaps or seaborn's "colorblind" palette, rather than a default red/green distinction as the only signal between two series.matplotlib/seaborn → static PNG. The default choice for a chart that's going into a report, an email, or anywhere it just needs to be an image. Save via plt.savefig("chart.png", dpi=150, bbox_inches="tight") — the bbox_inches="tight" avoids clipped axis labels, a common failure mode. Saved files surface to the user automatically as artifacts; you don't need to do anything extra to hand them over.plotly (+ kaleido for static export) → when the user wants interactivity (hover tooltips, zoom/pan) or explicitly asks for an HTML output. Export interactive output via fig.write_html("chart.html"); if a static image is needed instead, fig.write_image("chart.png") (uses kaleido under the hood).matplotlib/seaborn unless the user's request specifically implies interactivity or a web-embeddable artifact — it's the lighter-weight choice and covers the vast majority of "make me a chart" requests.Look at what you actually plotted against what was asked — a common failure mode is generating a technically-valid chart of the wrong slice of data (e.g. totals instead of averages, or the wrong grouping column) because a data-analysis step upstream computed something adjacent to, but not exactly, what was requested. Re-read the user's question once more against the chart's title and axes before presenting it.
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Build a new PipesHub connector, or extend an existing one, end to end, following CONNECTOR.md. Use when asked to add, create, build or port a connector or integration for an external source (Jira, Drive, Slack, a database, a SaaS API), or to add sync of a new object type to an existing connector.
The English summary is being prepared. Showing the original description.
Use this skill whenever the user asks you to load, clean, aggregate, join, or otherwise analyze tabular data (csv, xlsx, json, or a database export) — 'what's the total revenue by region', 'clean up this dataset', 'join these two files on customer id'. Establishes a verification discipline so reported numbers are always numbers the code actually printed, never numbers you inferred or estimated.
The English summary is being prepared. Showing the original description.
Use this skill whenever the user asks for a Word document, a .docx file, or a report/memo/letter/proposal delivered as one — creating a new document from scratch, editing an existing .docx, or reading/summarizing the contents of one. Covers page layout, headings, tables, numbered/bulleted lists, page breaks, and a table of contents. Does not cover converting a .docx to PDF or .doc to .docx (see file-conversion for what conversions ARE supported in this environment).
The English summary is being prepared. Showing the original description.
Use this skill when the user asks to convert a file from one format to another (e.g. csv to xlsx, xlsx to json, pdf text to csv) and you're not sure which library handles that specific pair, or whether it's supported in this environment at all. Dispatches to the right library per format pair and explicitly lists conversions that are NOT supported here so you don't attempt one that will silently fail or produce a wrong result.
The English summary is being prepared. Showing the original description.
Low-level support skill for hand-editing the raw XML inside a .docx or .pptx file (both formats are just ZIP archives of XML parts). Use this directly only when you need to unpack, inspect, or repack Open XML yourself — for example fixing a validation error in generated markup, or making one small targeted edit without regenerating a whole document from scratch. The docx and pptx skills already use this internally for their 'edit an existing file' workflow, so you rarely need to load this on its own.
The English summary is being prepared. Showing the original description.
Use this skill whenever the user asks for a PDF — generating a new one from scratch (reports, invoices, letters), extracting text/tables from an existing PDF, or filling in an existing PDF form. Does not cover converting an existing Word/PowerPoint/Excel file to PDF (that requires LibreOffice, not available in this environment — see file-conversion).
The English summary is being prepared. Showing the original description.