data-analysis-iteration
CommunityEnd-to-end on-device data analysis with AI agents.
System Documentation
What problem does it solve?
This skill enables end-to-end data-analysis on Windows ARM64 using NPU-accelerated models. It supports batch processing, OCR, vector search, and knowledge-graph management across eight agents to deliver actionable insights while keeping data on-device.
Core Features & Use Cases
- Eight-Agent collaboration orchestrates data flow from ingestion to final report.
- On-device NPU-accelerated inference ensures data privacy and low latency.
- Supports multi-format data (CSV/JSON/Excel) and OCR for images; enables batch processing and automated card generation.
- Builds and queries a local knowledge graph with vector search for contextual insights.
- Automated report generation and knowledge retention for future reuse.
Quick Start
Run batch_process.py on your data directory to preprocess files, trigger the eight-agent workflow, and generate a set of blue/green/yellow/red cards along with a final report. Ensure QNN SDK, GenieAPIService, and required Python packages are installed and configured as described in the repository references. Then review the generated cards in the knowledge dashboard and export the final report.
Dependency Matrix
Required Modules
Components
💻 Claude Code Installation
Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.
Please help me install this Skill: Name: data-analysis-iteration Download link: https://github.com/anbeime/antinet/archive/main.zip#data-analysis-iteration Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
Agent Skills Search Helper
Install a tiny helper to your Agent, search and equip skill from 223,000+ vetted skills library on demand.