Skip to content

AI Dev Local

Welcome to AI Dev Local - a comprehensive AI lab for local development with various AI services and Model Context Protocol (MCP) integrations.

Overview

AI Dev Local provides a unified platform to run and manage multiple AI services locally, enabling developers to:

  • Observe and analyze LLM interactions with Langfuse
  • Build visual workflows with FlowiseAI
  • Chat with models through Open WebUI
  • Proxy multiple LLM providers with LiteLLM
  • Integrate development tools via MCP servers

Architecture

System Context Diagram

C4Context
    title AI Dev Local - System Context

    Person(dev, "AI Developer", "Develops AI applications and experiments with models")
    Person(researcher, "AI Researcher", "Conducts experiments and analyzes model performance")
    Person(devops, "DevOps Engineer", "Manages infrastructure and deployment pipelines")
    Person(pm, "Product Manager", "Tracks AI feature usage and performance metrics")

    System(aidl, "AI Dev Local", "Local AI development platform with integrated services")

    System_Ext(llm_providers, "LLM Providers", "OpenAI, Anthropic, Local models, etc.")
    System_Ext(gitlab, "GitLab", "Source code management and CI/CD")
    System_Ext(github, "GitHub", "Source code management and collaboration")
    System_Ext(sonarqube, "SonarQube", "Code quality and security analysis")
    System_Ext(ide, "IDE/Editor", "VS Code, Cursor, Codium with MCP support")

    Rel(dev, aidl, "Uses CLI, builds workflows, chats with models")
    Rel(researcher, aidl, "Analyzes performance, tracks experiments")
    Rel(devops, aidl, "Manages services, monitors infrastructure")
    Rel(pm, aidl, "Reviews usage analytics, cost tracking")

    Rel(aidl, llm_providers, "Proxies requests, aggregates responses")
    Rel(aidl, gitlab, "Integrates via MCP for issues, CI/CD")
    Rel(aidl, github, "Integrates via MCP for code reviews")
    Rel(aidl, sonarqube, "Integrates via MCP for quality metrics")
    Rel(ide, aidl, "Connects to MCP servers")

Component Architecture

graph TB
    subgraph "Developer Laptop"
        CLI[AI Dev Local CLI]
        IDE[IDE/Editor<br/>💻 VS Code, Cursor, etc.]
        BROWSER[Web Browser<br/>🌐 Access to Web UIs]
    end

    subgraph "Web Services"
        LF[Langfuse<br/>📊 Observability]
        FW[FlowiseAI<br/>🎨 Workflow Builder]
        OW[Open WebUI<br/>💬 Chat Interface]
        LL[LiteLLM Proxy<br/>🚀 API Gateway]
        DOCS[Documentation<br/>📖 MkDocs Server]
    end

    subgraph "MCP Servers"
        GL[GitLab MCP<br/>📋 Issues & CI/CD]
        GH[GitHub MCP<br/>🔀 Code Reviews]
        SQ[SonarQube MCP<br/>🔍 Quality Analysis]
    end

    subgraph "External Integrations"
        GLR[GitLab API]
        GHR[GitHub API]
        SQR[SonarQube API]
        LLMP[LLM Providers<br/>OpenAI, Anthropic, etc.]
    end

    CLI --> LF
    CLI --> FW
    CLI --> OW
    CLI --> LL
    CLI --> DOCS

    CLI --> GL
    CLI --> GH
    CLI --> SQ

    BROWSER --> LF
    BROWSER --> FW
    BROWSER --> OW
    BROWSER --> DOCS

    GL --> GLR
    GH --> GHR
    SQ --> SQR

    FW --> LL
    OW --> LL
    LF --> LL
    LL --> LLMP

    IDE -.-> GL
    IDE -.-> GH
    IDE -.-> SQ

    style CLI fill:#e1f5fe
    style IDE fill:#e3f2fd
    style BROWSER fill:#f1f8e9
    style LF fill:#f3e5f5
    style FW fill:#e8f5e8
    style OW fill:#fff3e0
    style LL fill:#fce4ec
    style DOCS fill:#e8eaf6

Quick Start

Get started with AI Dev Local in minutes:

# Install with pipx
pipx install ai-dev-local

# Start all services
ai-dev-local start

# Check status
ai-dev-local status

Features

🔍 Observability

  • Track LLM usage and performance with Langfuse
  • Monitor costs and latency across providers
  • Debug and optimize AI workflows

🎨 Visual Workflows

  • Build AI workflows with drag-and-drop interface
  • Connect multiple AI services and APIs
  • Test and iterate on complex AI pipelines

💬 Chat Interface

  • Modern web interface for chatting with AI models
  • Support for multiple model providers
  • File uploads and conversation management

🚀 API Gateway

  • Unified API for multiple LLM providers
  • Load balancing and rate limiting
  • Cost tracking and usage analytics

🔧 Development Integration

  • GitLab integration for issue tracking and CI/CD
  • GitHub integration for code reviews and discussions
  • SonarQube integration for code quality analysis
  • IDE-side MCP servers for direct integration with VS Code, Codium, Cursor, and other editors

Getting Started

Ready to dive in? Choose your setup approach:

Quick Start Options

Configuration & Troubleshooting

MCP Integration Approaches

  1. IDE-side Integration - Run MCP servers on-demand within your editor
  2. Docker Compose Stack - Full containerized MCP server deployment
  3. Hybrid Approach - Combine both for maximum flexibility

Version

Current version: v0.2.1

Last updated: July 27, 2025