# npdeehan/agentic-patterns-bpmn architecture

> Architecture diagram of the GitHub repository npdeehan/agentic-patterns-bpmn, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2025-04-02). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/npdeehan/agentic-patterns-bpmn
- Repository: https://github.com/npdeehan/agentic-patterns-bpmn

## Overview

##### Main Components
- AI Agent Execution Module – This is the core “engine” that drives the AI processes.
- Safeguard/Intervention Processes – Each BPMN file (such as “Show Visuals”, “Increase Trust”, “Human in the Loop”, “Making Good Choices”, “Trigger Human Intervention”, “First Process for Doctor” and “Medical Intervention AI”) represents process checkpoints or decision branches within the safeguard architecture.
- Human Intervention Layer – This highlights where the system brings a human into the loop (clearly depicted in files with names like “Human in the Loop” and “Trigger Human Intervention”).

##### Relationships and Interactions
- Decision points that branch out into either automated execution or conditions requiring human intervention.
- Specifically, use components like “First Process for Doctor” and “Medical Intervention AI” to indicate advanced or specialized intervention scenarios.

##### Architectural Patterns and Design Principles
- Progressive escalation: each BPMN process represents an additional layer of scrutiny or safeguard.
- Decision nodes where the system can either continue execution or route the process to human oversight.
- Be sure to illustrate how control flows are managed and how feedback loops (like triggering human intervention after a decision point) are integrated into the system.

##### Incorporate Relevant Technologies or Frameworks
- Even though the project is not a full-stack application, any framework or notation used by BPMN should be noted.
- If any execution or simulation engine is implied for the process flows, include it as an “orchestrator” component.

## Diagram (Mermaid)

```mermaid
flowchart TD
    A["AI Agent Execution Module"]:::automated
    S["Show Visuals"]:::automated
    I["Increase Trust"]:::automated
    M["Making Good Choices"]:::automated
    D["Escalation?"]:::decision
    C["Continue Execution"]:::automated
    H["Human in the Loop"]:::human
    T["Trigger Human Intervention"]:::human
    F["First Process for Doctor"]:::specialized
    MI["Medical Intervention AI"]:::specialized

    A -->|"starts"| S
    S -->|"proceeds"| I
    I -->|"then"| M
    M -->|"decision"| D
    D -- "No" --> C
    C -->|"loops back"| A
    D -- "Yes" --> H
    H -->|"initiates"| T
    T -->|"specializes"| F
    F -->|"further action"| MI

    click A "https://github.com/npdeehan/agentic-patterns-bpmn/blob/main/README.md"
    click S "https://github.com/npdeehan/agentic-patterns-bpmn/blob/main/Show Visuals.bpmn"
    click I "https://github.com/npdeehan/agentic-patterns-bpmn/blob/main/Increase Trust.bpmn"
    click M "https://github.com/npdeehan/agentic-patterns-bpmn/blob/main/Making Good Choices.bpmn"
    click H "https://github.com/npdeehan/agentic-patterns-bpmn/blob/main/Human in the Loop.bpmn"
    click T "https://github.com/npdeehan/agentic-patterns-bpmn/blob/main/Trigger Human Intervention.bpmn"
    click F "https://github.com/npdeehan/agentic-patterns-bpmn/blob/main/First Process for Doctor.bpmn"
    click MI "https://github.com/npdeehan/agentic-patterns-bpmn/blob/main/Medical Intervention AI.bpmn"

    classDef automated fill:#b3e5fc,stroke:#01579b,stroke-width:2px;
    classDef human fill:#ffcdd2,stroke:#b71c1c,stroke-width:2px;
    classDef specialized fill:#c8e6c9,stroke:#2e7d32,stroke-width:2px;
    classDef decision fill:#fff9c4,stroke:#fbc02d,stroke-width:2px;
```
