agentleFS
santoshkanthetysantoshkanthety/powerbi-agent/skills/powerbi-python-visuals/SKILL.md

powerbi-python-visuals

Skill3 starsChanged 7 months agoMITView on GitHub

Python visual creation and matplotlib/seaborn patterns for PBIR reports. Automatically invoke when the user mentions "Python visual", "matplotlib in Power BI", "seaborn in Power BI", "pythonVisual", or asks to "create a Python visual", "add a matplotlib chart", "write a Python visual script".

  • Installs packages

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Create the visual.json file manually (see pbir-format skill in the pbip plugin for JSON structure) with visualType: pythonVisual, field …

    line 29
  2. Very specific:

    Only the last plt.show() call renders.

    line 48
  3. Very specific:

    Before presenting the script to the user, dispatch the python-reviewer agent to validate correctness and provide design feedback.

    line 52

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Python Visuals in Power BI (PBIR)

Report modification requires tooling. Two paths exist:

  1. pbir CLI (preferred) -- use the pbir command and the pbir-cli skill. Install with uv tool install pbir-cli or pip install pbir-cli. Check availability with pbir --version.
  2. Direct JSON modification -- if pbir is not available, use the pbir-format skill (pbip plugin) for PBIR JSON structure and patterns. Validate every change with jq empty <file.json>.

If neither the pbir-cli skill nor the pbir-format skill is loaded, ask the user to install the appropriate plugin before proceeding with report modifications.

Python visuals execute matplotlib/seaborn scripts to render static PNG images on the Power BI canvas. Prefer seaborn over raw matplotlib for cleaner syntax and better defaults -- it handles most chart types with less code.

Visual Identity

  • visualType: pythonVisual
  • Data role: Values (columns and measures, multiple allowed)
  • Data variable: dataset (pandas DataFrame, auto-injected)
  • Row limit: 150,000 rows
  • Output: Static PNG at 72 DPI -- no interactivity

Workflow: Creating a Python Visual

Step 1: Add the Visual

Create the visual.json file manually (see pbir-format skill in the pbip plugin for JSON structure) with visualType: pythonVisual, field bindings for the columns and measures you need (use Values:Table.Column or Values:Table.Measure format), and position/size as required.

Step 2: Write the Script

import matplotlib.pyplot as plt

fig, ax = plt.subplots(figsize=(8, 4))
ax.bar(dataset["Date"], dataset["Sales"], color="#5B8DBE")
ax.spines["top"].set_visible(False)
ax.spines["right"].set_visible(False)
plt.tight_layout()
plt.show()  # MANDATORY

Critical rules:

  • plt.show() is mandatory as the final line -- nothing renders without it
  • dataset is auto-injected as a pandas DataFrame; do not create it
  • Column names match the nativeQueryRef (display name) from field bindings
  • Only the last plt.show() call renders; multiple figures not supported

Step 2b: Review

Before presenting the script to the user, dispatch the python-reviewer agent to validate correctness and provide design feedback.

Step 3: Inject the Script

Set the script content in the visual's objects.script[0].properties.source literal value (see PBIR Format section below).

Escaping rules for visual.json injection:

The script must be encoded as a single-quoted DAX literal string inside expr.Literal.Value:

  • Newlines in the script become \n in the JSON string
  • Double quotes inside the script (e.g., "#5B8DBE") become \" in the JSON string
  • The entire script is wrapped in single quotes: 'import matplotlib...\nplt.show()'
  • See examples/visual/ for a complete real-world visual.json showing this encoding

Step 4: Validate

Validate JSON syntax with jq empty <visual.json> and inspect the visual.json to confirm script content and field bindings.

PBIR Format

Scripts are stored in visual.objects.script[0].properties:

{
  "source": {"expr": {"Literal": {"Value": "'import matplotlib.pyplot as plt\\n...\\nplt.show()'"}}},
  "provider": {"expr": {"Literal": {"Value": "'Python'"}}}
}

The CLI handles all escaping automatically.

Supported Libraries

Power BI Service (Python 3.11)

PackageVersionPurpose
matplotlib3.8.4Primary plotting
seaborn0.13.2Statistical visualization
numpy2.0.0Numerical computing
pandas2.2.2Data manipulation
scipy1.13.1Scientific computing
scikit-learn1.5.0Machine learning
statsmodels0.14.2Statistical models
pillow10.4.0Image processing

Not supported: plotly, bokeh, altair (networking blocked in Service).

Full package list: https://learn.microsoft.com/power-bi/connect-data/service-python-packages-support

Desktop

Any locally installed package works without restriction.

Best Practices

  1. Always call plt.show() -- mandatory, must be the final line
  2. Use figsize=(w, h) to match container aspect ratio (72 DPI output)
  3. Remove chart chrome -- ax.spines["top"].set_visible(False) etc.
  4. Use hex colors matching the report theme
  5. Keep scripts simple -- 5-min timeout Desktop, 1-min Service
  6. Minimize transforms -- do heavy computation in DAX/Power Query instead
  7. Use try/except for robustness in production scripts
  8. Copy data first -- data = dataset.copy() before manipulation

Limitations

ConstraintDesktopService
OutputStatic PNG, 72 DPIStatic PNG, 72 DPI
Timeout5 minutes1 minute
Row limit150,000150,000
Payload--30 MB
NetworkingUnrestrictedBlocked
GatewayPersonal onlyPersonal only
Cross-filter FROMNot supportedNot supported
Receive cross-filterYesYes
Publish to webNot supportedNot supported
Embed (app-owns-data)Ending May 2026Ending May 2026

Script Structure Template

import matplotlib.pyplot as plt
import numpy as np

# 1. Guard against empty data
if dataset.empty:
    fig, ax = plt.subplots(1, 1, figsize=(6, 4))
    ax.text(0.5, 0.5, "No data available", ha='center', va='center', fontsize=14, color='#888888')
    ax.axis('off')
    plt.show()
else:
    # 2. Data preparation (dataset is auto-injected)
    data = dataset.copy()

    # 3. Create figure with explicit size
    fig, ax = plt.subplots(figsize=(8, 4))

    # 4. Plot
    ax.plot(data["X"], data["Y"], color="#5B8DBE", linewidth=2)

    # 5. Style
    ax.spines["top"].set_visible(False)
    ax.spines["right"].set_visible(False)
    ax.grid(axis="y", alpha=0.3)

    # 6. Layout and render
    plt.tight_layout()
    plt.show()

When to Use Python Visuals

Python visuals are appropriate for statistical and analytical visualizations where the focus is on data analysis rather than interactivity. Use Python visuals when you need:

  • Statistical charts (distributions, regressions, correlations, violin plots)
  • Analytical visualizations leveraging scipy, scikit-learn, or statsmodels
  • Chart types that seaborn/matplotlib handle well but Power BI natives don't support

Output is static PNG -- no cross-filtering FROM the visual, no hover/tooltip interactivity. Use Deneb instead for interactive custom visuals. Use SVG measures for simple inline graphics in tables/cards.

References

  • references/community-examples.md -- seaborn gallery examples organized by chart type, plus matplotlib and Python Graph Gallery links
  • references/chart-patterns.md -- Common matplotlib/seaborn chart patterns (bar, heatmap, donut, KPI, area)
  • examples/script/ -- Standalone Python scripts (bar-chart, trend-line) -- ready to inject into visual.json after escaping
  • examples/visual/bar-chart.json -- PBIR visual.json: horizontal stacked bar with PY comparison lines and % change labels
  • examples/visual/kpi-card.json -- PBIR visual.json: text-based KPI with value, % change indicator, and PY comparison
  • examples/visual/trend-line.json -- PBIR visual.json: area chart with line plot and monthly x-axis

Fetching Docs

To retrieve current Python visual / package support docs, use microsoft_docs_search + microsoft_docs_fetch (MCP) if available, otherwise mslearn search + mslearn fetch (CLI). Search based on the user's request and run multiple searches as needed to ensure sufficient context before proceeding.

Related Skills

  • pbi-report-design -- Layout and design best practices
  • r-visuals -- R Script visuals (same concept, different language)
  • deneb-visuals -- Vega/Vega-Lite visuals (interactive, vector-based alternative)
  • svg-visuals -- SVG via DAX measures (lightweight inline graphics)
  • pbir-format (pbip plugin) -- PBIR JSON format reference

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.