| name | Chart Visualization |
| description | Render research-style PNG charts for trends, comparisons, distributions, and scatter relationships when a visual materially improves the answer |
| when_to_use | Use when the user asks for a chart, plot, visualization, trend line, bar chart, histogram, scatter plot, or when a numeric answer is easier to understand as a visual |
| aliases | ["chart visualization","chart","plot","visualize","trend chart"] |
| allowed-tools | ["skill_tool"] |
| user-invocable | true |
| context | inline |
| arguments | ["spec_json"] |
| script | scripts/render_chart.py |
| shell | python3 |
Chart Visualization
Use this skill when a chart will materially improve the answer. Do not render a chart just because it is possible.
Invocation
On native Codex, resolve this skill's directory from the disclosed SKILL.md
path and run the bundled script directly:
python3 <skill-dir>/scripts/render_chart.py '<JSON chart spec>'
Parse the script's JSON stdout and use the returned artifacts path. On a
legacy Hone runner without native skill execution, call:
skill_tool(
skill_name="chart_visualization",
execute_script=true,
script_arguments={"spec_json":"<JSON chart spec>"}
)
The script expects one JSON object string as spec_json.
Required Spec Fields
Optional Spec Fields
subtitle
x_label
y_label
x_values
annotations
footnotes
palette
output_name
Supported Chart Types
line
area
bar
scatter
histogram
horizontal_bar
Series Shape
Each series item should usually be:
{
"name": "Revenue",
"values": [100, 120, 135]
}
You may optionally add a .