insight_nl2code

Execute user-provided pandas code in a secure sandbox with serialized results.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/huangxn27/broadband-agent-demo --skill insight-nl2code
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: insight_nl2code
Source: https://github.com/huangxn27/broadband-agent-demo/tree/main/backend/skills/insight_nl2code
Command: npx skills add https://github.com/huangxn27/broadband-agent-demo --skill insight-nl2code

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires ce_insight_core, json_repair, and includes scripts (resource) and references (resource) components.

What problem does it solve?

NL2Code sandbox enables executing user-provided pandas code within a secure sandbox to perform bespoke analyses that are not covered by the standard 12 insight functions.

Core Features & Use Cases

  • Sandbox execution of pandas code with safety constraints and structured output
  • Supports advanced queries like Top N, custom multi-column comparisons, and derived metrics
  • Returns serialized results in a consistent format for downstream tooling or dashboards

Quick Start

Provide a payload with code and query_config to the NL2Code sandbox and observe the structured result.

Frequently Asked Questions about insight_nl2code

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run custom pandas code for data analysis beyond standard insights?

To run custom pandas code for data analysis, you can execute user-provided scripts in a secure sandbox. This enables bespoke data transformations like Top N selections and multi-column comparisons beyond built-in insight functions.

Can I perform Top N selections and multi-column comparisons using pandas safely?

Yes, you can perform Top N selections and multi-column comparisons safely. The sandbox environment enforces restricted execution with explicit input and output contracts, ensuring custom pandas operations remain reproducible.

How do I execute bespoke data transformations in a restricted pandas environment?

You execute bespoke data transformations by providing a payload with custom code and a query configuration. The sandbox processes the pandas operations and returns serialized results in a consistent format for downstream tooling.

Does the custom pandas sandbox support structured output for downstream dashboards?

Yes, the custom pandas sandbox supports structured output for downstream dashboards. It returns serialized results in a consistent format, ensuring that outputs from your bespoke data analysis are readily available for integration.

What are the limitations of using a sandbox for custom pandas data analysis?

The sandbox approach for custom pandas data analysis enforces a restricted environment with safety constraints. While it supports advanced queries, operations must adhere to explicit input and output contracts to ensure execution safety and reproducibility.