README.md

Data Analysis Loop

Status: Planned

For data cleaning, exploratory analysis, visualization, insight generation, and report creation.

Flow

Load Data → Validate Schema → Clean Data → Explore → Analyze → Visualize → Insight Review → Final Report

Roles

Data Loader

• Load data from files, databases, or APIs.

Data Cleaner

• Validate schema, handle missing values, fix inconsistencies.

Analyst

• Perform statistical analysis, hypothesis testing, and feature exploration.

Visualization Agent

• Generate charts and dashboards for key findings.

Insight Reviewer

• Review findings for accuracy and business relevance.

Human Checkpoints

• Confirm analysis goal
• Review final insights

Source / Attribution

• Type: inspired_by
• Name: CrewAI Stock Analysis / LangGraph RAG and analysis examples
• URL: https://github.com/crewAIInc/crewAI-examples, https://github.com/langchain-ai/langgraph/tree/main/examples
• License: please verify
• Notes: Inspired by public agent analysis examples. This loop should be platform-agnostic and not copy framework-specific code.

TODO

• Write template/ORCHESTRATOR.md
• Write template/state.template.md
• Write prompts for each role
• Add heartbeat configuration
• Add failure behavior