Skip to main content

AADC - Agentic AI Developer Console. Build anything with AI.

Project description

๐Ÿ”ด AADC - Agentic AI Developer Console

An AI-powered CLI tool that creates complete, production-ready applications from natural language instructions. Built with Google's Gemini API.

Version Python

โœจ Features

  • Multi-Model Support: Switch between Gemini, OpenAI GPT, and Claude models
  • Project Analysis: /init analyzes your codebase for AI context
  • Complete Code Generation: Creates fully functional websites, apps, and games from a single instruction
  • File Management: Create, read, write, edit, and delete files and folders
  • Background Terminals: Run servers (Node.js, Python, etc.) in background sessions
  • Persistent Memory: Remembers your preferences and project details across sessions
  • Permission Modes: Control when the AI asks for confirmation
  • Plan Mode: Review AI's plan before it starts coding
  • Command Execution: Run shell commands, install packages, start servers
  • Iterative Development: Refine and modify your projects through conversation

๐Ÿš€ Quick Start

1. Installation

# Clone or download this repository
cd aadc

# Install dependencies
pip install -r requirements.txt

2. Get Your API Key

  1. Go to Google AI Studio
  2. Create or sign in with your Google account
  3. Generate an API key

3. Set Your API Key

Option A: .env File (Recommended)

# Copy the example file
cp .env.example .env

# Edit .env and add your key
GEMINI_API_KEY=your-api-key-here

Option B: Environment Variable

export GEMINI_API_KEY="your-api-key-here"

Option C: Enter When Prompted The tool will ask for your API key on first run and optionally save it.

4. Run the Agent

python main.py

๐Ÿ“– Usage

Interactive Mode

Simply run the agent and start describing what you want to build:

$ python main.py

โ•”โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•—
โ•‘                    GEMINI CODING AGENT                       โ•‘
โ•šโ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•โ•

โ”Œโ”€[/home/user/projects]
โ””โ”€โ–ถ Create a snake game with HTML, CSS, and JavaScript

๐Ÿ”ง Tool: create_folder
   โ”œโ”€ folder_path: snake-game
   โ””โ”€ โœ“ Success

๐Ÿ”ง Tool: create_file
   โ”œโ”€ file_path: snake-game/index.html
   โ”œโ”€ content: <!DOCTYPE html>...
   โ””โ”€ โœ“ Success

...

๐Ÿค– Assistant:
I've created a complete Snake game! To play:
1. Open snake-game/index.html in your browser
2. Use arrow keys to control the snake
3. Eat the red food to grow and score points

Single Command Mode

Run a single command without entering interactive mode:

python main.py "Create a todo app with React"

CLI Options

python main.py --help

Options:
  prompt              Optional prompt for single command mode
  -k, --api-key       Gemini API key (overrides env variable)
  -d, --directory     Working directory for file operations
  -v, --version       Show version

Built-in Commands

Command Description
/init Analyze codebase, generate AI context summary
/models [name] List or switch AI models
/help Show help information
/perm [mode] View/set permission mode (ask/auto/command_ask)
/plan Toggle plan mode on/off
/terminals List active background terminal sessions
/kill <id> Stop a background terminal
/output <id> View output from a terminal
/memory List all stored memories
/mem <query> Search memories
/forget <key> Remove a memory
/clear Clear conversation history
/cd <path> Change working directory
/pwd Print current directory
/ls List files in current directory
/exit Exit the agent

Permission Modes

Use /perm <mode> to control when the AI asks for confirmation:

Mode Description
ask Ask before executing ANY tool
auto Execute all tools automatically (default)
command_ask Only ask before shell commands

Plan Mode

Enable plan mode to have the AI create a plan.md first:

  • Use /plan to toggle on/off
  • Or prefix your request with plan: (e.g., "plan: create a todo app")
  • Review the plan and approve/disapprove before coding begins

Available Models

Use /models <name> to switch between AI providers:

Model Provider Description
gemini-3-pro-preview Google Gemini 3 Pro Preview (default)
gemini-2.5-pro Google Gemini 2.5 Pro
GPT-5.1 OpenAI GPT-5.1
gpt-5-pro OpenAI GPT-5 Pro
claude-sonnet-4-5 Anthropic Claude Sonnet 4.5
claude-opus-4-5 Anthropic Claude Opus 4.5

Project Initialization

Run /init to analyze your codebase:

  • Generates .aadc-summary.md with project overview
  • AI uses this for better context understanding
  • Run again after major changes to refresh

CLI Options

python main.py --help

Options:
  prompt              Optional prompt for single command mode
  -k, --api-key       Gemini API key (overrides env variable)
  -d, --directory     Working directory for file operations
  -p, --plan          Start in plan mode
  -m, --mode          Permission mode (ask/auto/command_ask)
  -v, --version       Show version

๐ŸŽฏ Example Prompts

Web Development

Create a responsive portfolio website with a dark theme, about section, 
project gallery, and contact form

Games

Build a Tetris clone with score tracking and increasing difficulty

Backend

Create a REST API with Express.js for a blog with posts and comments, 
including MongoDB integration

Utilities

Build a Python CLI tool that converts images to different formats 
with support for batch processing

Full-Stack

Create a real-time chat application with React frontend and 
Socket.io backend

๐Ÿ› ๏ธ Available Tools

The agent has access to 19 tools organized into three categories:

File Operations

Tool Description
create_file Create a new file with content
create_folder Create a directory
delete_file Delete a file
delete_folder Delete a directory recursively
list_files List directory contents
read_file Read file content
write_file Write/overwrite file content
edit_file Make surgical edits to existing files
execute_command Run shell commands

Terminal Management

Tool Description
open_terminal Start a background process (servers, etc.)
close_terminal Stop a background terminal
list_terminals List all active terminals
get_terminal_output Get output from a terminal
send_terminal_input Send input to a terminal

Memory (Persistent)

Tool Description
remember Store information for future sessions
recall Retrieve a specific memory
search_memory Search through memories
list_memories List all stored memories
forget Remove a memory

๐Ÿ”ง Configuration

Default settings can be modified in config.py:

DEFAULT_CONFIG = {
    "model": "gemini-2.5-pro-preview-05-06",
    "max_tokens": 65536,
    "temperature": 0.7,
    "timeout": 300,
    "max_iterations": 50,
}

๐Ÿ“ Project Structure

aadc/
โ”œโ”€โ”€ main.py              # CLI entry point
โ”œโ”€โ”€ agent.py             # Multi-provider AI agent (Gemini/OpenAI/Anthropic)
โ”œโ”€โ”€ tools.py             # Tool implementations (19 tools)
โ”œโ”€โ”€ prompts.py           # System prompt
โ”œโ”€โ”€ config.py            # Configuration & model management
โ”œโ”€โ”€ utils.py             # Display utilities
โ”œโ”€โ”€ terminal_manager.py  # Background terminal handling
โ”œโ”€โ”€ memory.py            # Persistent memory system
โ”œโ”€โ”€ project_init.py      # Codebase analysis & summarization
โ”œโ”€โ”€ requirements.txt     # Dependencies
โ”œโ”€โ”€ .env.example         # API key template
โ””โ”€โ”€ README.md            # This file

๐Ÿ”’ Security Notes

  • API keys are stored in ~/.aadc/.env with restricted permissions
  • The agent can execute shell commands - use with appropriate caution
  • Review generated code before running in production environments

๐Ÿ› Troubleshooting

"No API key found"

Set your API key via environment variable or enter it when prompted.

"Model not found"

Ensure you have access to the Gemini API and the model name is correct.

"Tool execution failed"

Check file permissions and ensure paths are valid.

๐Ÿ“ License

MIT License - feel free to use and modify as needed.

๐Ÿค Contributing

Contributions welcome! Feel free to submit issues and pull requests.


Built with โค๏ธ using Google Gemini API

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

aadc_cli-1.0.2.tar.gz (70.1 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

aadc_cli-1.0.2-py3-none-any.whl (74.8 kB view details)

Uploaded Python 3

File details

Details for the file aadc_cli-1.0.2.tar.gz.

File metadata

  • Download URL: aadc_cli-1.0.2.tar.gz
  • Upload date:
  • Size: 70.1 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.0

File hashes

Hashes for aadc_cli-1.0.2.tar.gz
Algorithm Hash digest
SHA256 93f6fa0b58774ea3001807e30f0ec533fad81a68958275ef74148e5504b0a4e6
MD5 616fc489d0e97149f9f0e2ff37a40dea
BLAKE2b-256 ef5bacea8efd7a002c658b4fabc7d4ecf380d7ed8d5ad6862a6e72ffea4da99c

See more details on using hashes here.

File details

Details for the file aadc_cli-1.0.2-py3-none-any.whl.

File metadata

  • Download URL: aadc_cli-1.0.2-py3-none-any.whl
  • Upload date:
  • Size: 74.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.0

File hashes

Hashes for aadc_cli-1.0.2-py3-none-any.whl
Algorithm Hash digest
SHA256 39e40c2c81c4db3fa94a1dc06fc1c242f044bd450766bbafaeeaa1de8bc68016
MD5 dc54df999dada8b779ee37f45ab7f8ec
BLAKE2b-256 f2306d1f591025d897445c09e47c40b57b0497bf8fbfb23d32d1dba42013bfdf

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page