deep-analytics

Generate self-contained ECharts HTML analytics pages from BigQuery data.

1|Updated Apr 7, 2026
One-click install
npx skills add https://github.com/al-stev/field-eng-skills --skill deep-analytics
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: deep-analytics
Source: https://github.com/al-stev/field-eng-skills/tree/main/.claude/skills/deep-analytics
Command: npx skills add https://github.com/al-stev/field-eng-skills --skill deep-analytics

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires pandas, numpy, google-cloud-bigquery, and includes scripts (resource) components.

What problem does it solve?

Deep Analytics automates the generation of shareable, self-contained HTML analytics pages from BigQuery so Solutions Engineers and analysts can quickly surface per-user and per-team intelligence without building bespoke reports or dashboards.

Core Features & Use Cases

  • Multi-dimensional pages: user journeys, cohort analysis, engagement decay, feature velocity, team detection, risk scoring, usage correlation, SDK versions, and performance deep dives.
  • BigQuery-driven: runs SQL queries, validates schemas, checks data availability, and transforms results into ECharts-powered visualizations and narrative sections.
  • Production-ready outputs: produces single-file HTML reports with inline data and themes suitable for sharing in customer conversations and technical reviews.
  • Privacy & safety: enforces cohort minimums and scans outputs to prevent SFDC account ID leakage, with descoping gates when required data or schemas are missing.

Quick Start

Run the deep analytics pipeline for a customer and page type to produce a self-contained HTML analytics page.

Frequently Asked Questions about deep-analytics

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

FAQPage Schema
How do I generate interactive HTML analytics reports from BigQuery data?

The pipeline generates interactive HTML analytics reports by running SQL queries against BigQuery, validating schemas, and transforming results into self-contained ECharts-powered visualizations. It checks data availability and outputs single-file reports ready for sharing.

What user journey and cohort analysis metrics can I extract from BigQuery for customer accounts?

From BigQuery, you can extract user journeys, cohort retention, engagement decay, feature velocity, team detection, risk scoring, usage correlation, SDK distributions, and performance signals. These metrics are transformed into multi-dimensional analytical pages for per-customer intelligence.

Do I need Application Default Credentials to access BigQuery for generating analytics pages?

Yes, Application Default Credentials are required for BigQuery access. The analytics pipeline enforces ADC-based authentication and requires a customer-account mapping configuration in a YAML file to generate the analytical pages successfully.

How does the pipeline handle data privacy and prevent account ID leakage in HTML reports?

To handle data privacy, the pipeline enforces cohort minimums and scans outputs to prevent SFDC account ID leakage. It also implements descoping gates that halt report generation when required data or schemas are missing.

Can I use pandas and numpy with BigQuery results to build ECharts visualizations?

Yes, pandas and numpy are used to transform BigQuery query results into ECharts-powered visualizations. These libraries handle the data manipulation required to produce the narrative sections and graphical outputs in the final self-contained HTML file.

What happens when required schemas or data are missing during the BigQuery analytics generation?

When required schemas or data are missing, the pipeline triggers descoping gates that prevent the analytics page from generating. It validates schema and data availability upfront to ensure accurate cohort analysis, user journey reporting, and risk scoring.