Astron Agent Project Modules

Project Overview

Astron Agent is an enterprise-grade, commercially-friendly Agentic Workflow development platform that integrates AI workflow orchestration, model management, AI & MCP tools, RPA automation, and team collaboration features.

Architecture Diagram

Module List

UI Layer

1. Console Frontend

Module Path: console/frontend/

Language: TypeScript + React

Main Responsibilities:

• Provide web user interface (SPA - Single Page Application)
• Agent creation and configuration UI
• Visual workflow editor
• Knowledge base management interface
• Model management and configuration
• Real-time chat window
• Multi-tenant space management

Tech Stack: React 18, TypeScript 5, Vite 5, Ant Design 5, Tailwind CSS, ReactFlow, Recoil/Zustand

Console Backend Layer

2. Console Backend

Module Path: console/backend/

Language: Java

Main Responsibilities:

• Provide REST API and SSE interfaces for management console
• User authentication and permission management
• CRUD interfaces for Agent, Workflow, and Knowledge
• Model management and configuration APIs
• File upload/download services
• Data statistics and analytics

Tech Stack: Spring Boot 3.5.4, MyBatis Plus 3.5.7, Spring Security, OAuth2

Sub-modules:

• hub: Main API service module
• toolkit: Utility module
• commons: Common module (DTOs, utilities, etc.)

Core Microservices Layer

3. Agent Service

Module Path: core/agent/

Language: Python

Main Responsibilities:

• Agent core execution engine
• Support multiple Agent types (Chat Agent, CoT Agent, CoT Process Agent)
• Agent lifecycle management
• Tool invocation and plugin integration
• Session management and context persistence

Tech Stack: FastAPI, SQLAlchemy 2.0, Pydantic, OpenTelemetry

Architecture Design: Follows DDD (Domain-Driven Design) with API layer, service layer, domain layer, and repository layer

4. Workflow Service

Module Path: core/workflow/

Language: Python

Main Responsibilities:

• Workflow orchestration and execution engine (Spark Flow)
• Multi-step process automation
• Workflow version management
• Event-driven asynchronous processing
• Visual workflow runtime debugging

Tech Stack: FastAPI, SQLModel, SQLAlchemy 2.0, Kafka (event streaming), LangChain

Event Mechanism: Event communication via Kafka Topic workflow-events

5. Knowledge Service

Module Path: core/knowledge/

Language: Python

Main Responsibilities:

• Knowledge base management and document processing
• Document vectorization and semantic search
• RAG (Retrieval-Augmented Generation) implementation
• LLM integration and embeddings generation
• Support for multiple document format parsing

Tech Stack: FastAPI, RAGFlow SDK, OpenAI API, SQLModel, Redis

Event Mechanism: Event communication via Kafka Topic knowledge-events

6. Memory DB Service

Module Path: core/memory/

Language: Python

Main Responsibilities:

• Conversation history storage and retrieval
• Context management (long-term and short-term memory)
• Session data persistence

Tech Stack: Python, database abstraction layer

7. Tenant Service

Module Path: core/tenant/

Language: Go

Main Responsibilities:

• Multi-tenant management
• Space isolation and permission control
• Organization structure management
• Resource quota management

Tech Stack: Go 1.23, Gin framework, MySQL

Design Philosophy: Implemented in Go for high performance and low memory overhead

Plugin System

8. Plugin: AI Tools

Module Path: core/plugin/aitools/

Language: Python

Main Responsibilities:

• Integration with iFLYTEK AI tools (IFLYTEX API)
• Third-party AI tool integration
• Tool invocation management and result caching

Tech Stack: FastAPI, HTTP Client

9. Plugin: RPA

Module Path: core/plugin/rpa/

Language: Python

Main Responsibilities:

• RPA process automation
• Process recording and playback
• Automated script execution
• Integration with external RPA executors

Tech Stack: FastAPI, RPA SDK

10. Plugin: Link

Module Path: core/plugin/link/

Language: Python

Main Responsibilities:

• External link resource integration
• URL content fetching and processing
• Link validation and metadata extraction

Tech Stack: FastAPI, HTTP Client

Common Services Layer

11. Common Module

Module Path: core/common/

Language: Python

Main Responsibilities:

• Provide cross-project common services and utilities
• Authentication and audit system (MetrologyAuth)
• Observability support (OTLP, OpenTelemetry)
• Database, cache, and message queue connection management
• Unified logging system
• OSS (MinIO) object storage integration

Tech Stack: Python, SQLModel, Redis Client, Kafka Client, OpenTelemetry

Core Value: Provide unified infrastructure abstraction for all Python microservices

Infrastructure Components (Data Mgmt & Messaging)

Data Persistence

• MySQL: Primary database for structured data storage
• Redis: Cache service, session storage, event registry
• PostgreSQL: Optional auxiliary database

Message Queue

• Kafka: Event streaming and inter-service communication
  • Topic: workflow-events - Workflow events
  • Topic: knowledge-events - Knowledge events
  • Topic: agent-events - Agent events

Object Storage

• MinIO: File storage service (PUT/GET operations)

External Service Integrations

LLM Providers

• Integration with multiple large language model services (OpenAI, Azure OpenAI, local models, etc.)
• Unified LLM invocation interface

IFLYTEX API

• iFLYTEK AI tool API integration
• Invoked through AI Tools plugin

RPA Executors

• External RPA automation executors
• Task distribution and execution via RPA plugin

Module Dependencies

Hierarchical Structure

UI Layer
    └── Console Frontend (React/TS)
         ↓ HTTP/REST/SSE

Console Backend Layer
    └── Console Backend (Java Spring Boot)
         ↓ HTTP/REST

Core Microservices Layer
    ├── Agent Service (Python FastAPI)
    ├── Workflow Service (Python FastAPI)
    ├── Knowledge Service (Python FastAPI)
    ├── Memory DB Service (Python)
    ├── Tenant Service (Go Gin)
    ├── Plugin: AI Tools (Python FastAPI)
    ├── Plugin: Link (Python FastAPI)
    └── Plugin: RPA (Python FastAPI)
         ↓

Common Services Layer
    └── Common Module (Python)
         ↓

Data & Messaging Layer
    ├── MySQL (Relational Database)
    ├── Redis (Cache/Session)
    ├── Kafka (Event Streaming)
    └── MinIO (Object Storage)
         ↓

External Services
    ├── LLM Providers (Large Language Models)
    ├── IFLYTEX API (iFLYTEK API)
    └── RPA Executors (RPA Executors)

Communication Patterns

Module Dependency Matrix

Technology Stack Summary

Development Standards

Python Modules

• Architecture: DDD (Domain-Driven Design)
• Code Style: Black + isort
• Type Checking: MyPy
• Code Analysis: Pylint, Flake8
• Testing: Pytest (coverage ≥ 70%)

Java Modules

• Architecture: Spring Boot layered architecture
• Code Style: Checkstyle
• Code Analysis: PMD
• Testing: JUnit

TypeScript Modules

• Code Style: ESLint + Prettier
• Type Checking: TypeScript strict mode
• Testing: Jest + React Testing Library

Go Modules

• Code Style: Go fmt
• Code Analysis: Go vet, Golint

Related Documentation

• Project README
• Deployment Guide
• Configuration Guide
• Agent Development Guide
• Frontend Development Guide

Document Version: v1.0 Last Updated: 2025-11-25