Skip to content

Atum246/droidclaw

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

20 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🤖 DroidClaw — The Ultimate Autonomous AI Agent for Android

Not a chat app. A fully autonomous AI agent that can do ANYTHING on your phone.
It thinks, acts, builds, creates, automates, and even makes its own tools when it doesn't have one.


📥 Download

Latest Release

Architecture Download Size
📱 Universal APK ⬇️ Download ~50MB
🔵 ARM64 (most phones) ⬇️ Download ~35MB
🟢 ARMv7 (older phones) ⬇️ Download ~30MB
🟠 x86_64 (emulators) ⬇️ Download ~35MB

Install Steps

  1. Download the APK for your device architecture
  2. Enable "Install from unknown sources" in your phone settings
  3. Open the APK file to install
  4. Launch DroidClaw and go through the onboarding setup
  5. Add your AI provider API key (OpenAI, Anthropic, Google, etc.)
  6. Start commanding your agent! 🚀

Requirements: Android 5.0+ (API 21), ~100MB storage, internet connection for cloud AI models


🧠 What Is DroidClaw?

DroidClaw is a fully autonomous AI agent platform for Android — inspired by OpenClaw. Unlike regular chat apps that just talk back, DroidClaw thinks and acts:

  • 🔧 Executes tools — 90+ tools across 16 categories
  • 🤖 Spawns sub-agents — parallel workers for complex tasks
  • 🔧 Creates its own tools — if it doesn't have one, it builds one
  • Creates its own skills — auto-learns new task types
  • 🌐 Browser automation — controls web pages like a human
  • 📱 Full phone control — calls, SMS, apps, settings, camera
  • 🖥️ Remote control — SSH into laptops, control PCs
  • 🔬 Deep research — multi-step investigation on any topic
  • 💓 Proactive heartbeat — checks things without being asked
  • 🖼️ Multi-modal — understands images, voice, documents
  • 🎨 Workflow builder — visual automation (like n8n on your phone)
  • 📡 Streaming responses — words appear in real-time
  • 🧠 Persistent memory — remembers everything across sessions
  • 🎤 Voice I/O — speak to it, it speaks back
  • 📥 Local AI models — run AI directly on your phone, offline

DroidClaw doesn't just chat. It EXECUTES.


⚡ Quick Start

1. Install APK → 2. Open app → 3. Name yourself → 4. Set timezone
→ 5. Add API key → 6. Start talking → 7. Watch it work

The onboarding walks you through everything conversationally — just like meeting a new person.


🤖 Autonomous Agent Features

🤖 Sub-Agent Spawning

DroidClaw can spawn isolated sub-agents that run their own reasoning loops in parallel. For complex tasks, it automatically suggests splitting work across multiple agents.

You: "Research Flutter vs React Native and also check my calendar"
DroidClaw: *Spawns two sub-agents — one for research, one for calendar*

🔧 Auto-Tool Creation

When DroidClaw encounters a task it doesn't have a tool for, it creates one on the fly:

You: "Post to my Instagram"
DroidClaw: *No Instagram tool exists*
DroidClaw: *Analyzes the task, creates instagram_post tool*
DroidClaw: *Registers it dynamically, executes it*
DroidClaw: "✅ Instagram post created!"

⚡ Auto-Skill Creation

Same for skills — if you ask it to do something it doesn't have a skill for:

You: "Write me a grant proposal"
DroidClaw: *No grant writing skill exists*
DroidClaw: *Creates grant_writer skill with triggers, system prompt, category*
DroidClaw: *Uses it and remembers for next time*

🔬 Deep Research Engine

Multi-layer research that goes deep:

You: "Research the impact of AI on healthcare"
DroidClaw:
  🔍 Layer 1: Generates research questions
  📚 Layer 2: Searches + fetches top sources
  🧠 Layer 3: Generates follow-up questions
  📚 Layer 4: Deeper investigation
  📝 Final: Synthesizes comprehensive report with citations

💓 Proactive Heartbeat

DroidClaw checks things without being asked:

  • 📅 Upcoming calendar events
  • 🔋 Low battery warnings
  • 🌧️ Severe weather alerts
  • Configurable interval (default: 30 min)
  • Quiet hours (won't bother you at night)

🖼️ Multi-Modal Processing

Send it anything and it understands:

  • 📸 Images → describe, OCR text, analyze, identify objects
  • 🎤 Voice notes → transcribe, analyze, extract commands
  • 📄 Documents → read, summarize, extract data
  • Auto-detects file type and processes accordingly

🔧 90+ Tools — Complete List

📁 File System (8)

Tool Description
read_file Read file contents
write_file Write content to file
list_dir List directory contents
search_files Search for files by pattern
delete_file Delete file/folder
move_file Move/rename file
copy_file Copy file
file_info Get file details

🌐 Web (5)

Tool Description
web_search Search the web (DuckDuckGo)
fetch_url Fetch and extract URL content
browse_web Open URL in browser
download_file_url Download file from URL
screenshot_web Screenshot a webpage

🌐 Browser Automation (20)

Tool Description
browser_open Open URL in automated browser
browser_navigate Navigate to URL
browser_click Click element by selector/text
browser_type Type text into input field
browser_select Select dropdown option
browser_scroll Scroll page up/down
browser_fill_form Fill multiple form fields
browser_screenshot Capture page screenshot
browser_get_text Extract text from elements
browser_get_html Get page/element HTML
browser_get_url Get current page URL
browser_execute_js Execute JavaScript on page
browser_extract_links Extract all links as JSON
browser_extract_table Extract tables as JSON
browser_login Automated login flow
browser_search Search on any website
browser_wait Wait for elements to appear
browser_multi_step Run multi-step automation
browser_close Close browser
browser_download Download file from page

📞 Calls & Contacts (8)

Tool Description
make_call Make a phone call
end_call End current call
answer_call Answer incoming call
reject_call Reject incoming call
get_call_log Get call history
get_contacts Get contacts list
add_contact Add new contact
block_contact Block a number

💬 Messaging (7)

Tool Description
send_sms Send text message
read_sms Read SMS messages
send_email Send email
send_whatsapp Send WhatsApp message
send_telegram Send Telegram message
share_content Share to any app
broadcast_sms Send SMS to multiple contacts

⏰ Scheduling (8)

Tool Description
set_alarm Set alarm with days, label
delete_alarm Delete an alarm
get_alarms List all alarms
start_timer Start countdown timer
stop_timer Stop running timer
set_reminder Set a reminder
get_calendar Get calendar events
add_calendar Add calendar event

📱 Phone Control (20)

Tool Description
set_brightness Set screen brightness
get_brightness Get current brightness
set_volume Set volume level
get_volume Get volume level
toggle_wifi Toggle WiFi
get_wifi_info Get WiFi status
toggle_bluetooth Toggle Bluetooth
toggle_mobile_data Toggle mobile data
toggle_airplane_mode Toggle airplane mode
toggle_flashlight Toggle flashlight
toggle_hotspot Toggle WiFi hotspot
toggle_do_not_disturb Toggle DND mode
toggle_auto_rotate Toggle auto-rotate
set_screen_timeout Set screen off time
lock_screen Lock the screen
take_screenshot Take a screenshot
record_screen Record screen
set_wallpaper Set wallpaper
set_ringtone Set ringtone
vibrate Vibrate device

📱 App Management (6)

Tool Description
open_app Open any app
close_app Force close app
list_apps List installed apps
install_app Install APK
uninstall_app Uninstall app
app_info Get app details

📍 Location & Maps (5)

Tool Description
get_location Get GPS location
get_address Get address from coordinates
open_maps Open maps with directions
share_location Share live location
geofence Set location alert

📸 Camera & Media (8)

Tool Description
take_photo Take a photo
record_video Record video
get_gallery Get photos from gallery
delete_photo Delete a photo
scan_qr Scan QR/barcode
generate_qr Generate QR code
get_storage_info Get storage usage
screen_brightness_auto Auto brightness

🔊 Media Playback (5)

Tool Description
play_music Play audio file
pause_music Pause playback
stop_music Stop playback
next_track Next track
text_to_speech Speak text aloud

🔐 Security (4)

Tool Description
authenticate Biometric authentication
check_biometric Check biometric availability
encrypt_data Encrypt data
decrypt_data Decrypt data

📊 Device Info (6)

Tool Description
get_device_info Full device info
get_battery Battery level & status
get_network_info Network & IP info
get_running_processes Running processes
get_system_settings System settings
vibrate Vibrate device

🖥️ Remote Control (10)

Tool Description
remote_connect Connect to remote device (SSH)
remote_disconnect Disconnect
remote_shell Run command on remote
remote_file_read Read file from remote
remote_file_write Write file to remote
remote_file_list List remote files
remote_screenshot Screenshot remote screen
remote_keyboard Send keystrokes
remote_mouse Move/click mouse
remote_clipboard Get/set clipboard

💻 Code Execution (4)

Tool Description
run_code Execute code in sandbox
run_shell Run shell command
run_python Run Python script
run_js Run JavaScript

🧠 Memory (4)

Tool Description
remember Save to long-term memory
recall Search memory
forget Delete memory entry
list_memories List all memories

🔗 Integration (5)

Tool Description
http_request Make HTTP API call
create_note Create a note
create_todo Create todo item
translate_text Translate text
clipboard_copy Copy to clipboard

🤖 Meta: Sub-Agents & Auto-Creation (9)

Tool Description
spawn_agent Spawn sub-agent for parallel tasks
list_agents List all running sub-agents
kill_agent Kill a sub-agent
steer_agent Send message to sub-agent
deep_research Multi-step deep research
create_tool Auto-create a new tool
create_skill Auto-create a new skill
list_created_tools List auto-created tools
list_created_skills List auto-created skills

🖼️ Multi-Modal (4)

Tool Description
process_image Analyze/OCR/describe images
process_voice Transcribe/analyze voice notes
process_document Read/summarize documents
process_any Auto-detect and process any file

🎨 Workflow (5)

Tool Description
create_workflow Create automation workflow
create_workflow_nl Create workflow from description
run_workflow Run a workflow
list_workflows List all workflows
list_workflow_templates List workflow templates

🎮 External (3)

Tool Description
read_nfc Read NFC tag
write_nfc Write NFC tag
control_smart_home Control smart devices

Total: 90+ tools and growing!


⚡ 110+ Skills — Complete List

💻 Development (15)

Code Review, Code Explainer, Code Converter, Code Generator, Debugger, Refactorer, Test Generator, API Designer, SQL Generator, Regex Builder, Git Helper, Docker Helper, JSON Formatter, Code Optimizer, Error Decoder

✍️ Writing (15)

Email Writer, Essay Writer, Story Writer, Poet, Report Writer, Cover Letter, Resume Builder, Grammar Check, Summarizer, Translator (100+ languages), Paraphraser, Headline Generator, Blog Writer, Slogan Creator, Script Writer

📊 Analysis (10)

Data Analyzer, Sentiment Analysis, Market Research, SWOT Analysis, Risk Assessment, Comparator, Brainstormer, Critic, Researcher, Trend Analyzer

🎓 Education (10)

Math Solver, Science Explainer, Historian, Flashcard Maker, Quiz Generator, ELI5, Study Planner, Language Tutor, Concept Explainer, Vocabulary Builder

💼 Business (10)

Business Plan, Pitch Deck, Meeting Notes, Project Planner, Budget Helper, Contract Review, Strategy Advisor, Presentation Maker, Invoice Generator, KPI Tracker

🎨 Creative (10)

Logo Ideas, Color Palette, UI Advisor, Brand Namer, Recipe Creator, Travel Planner, Gift Advisor, Interior Design, Fashion Advisor, Music Theory

🛠️ Productivity (10)

Todo Manager, Calendar Helper, Note Taker, Decision Maker, Time Manager, Habit Tracker, Goal Setter, Focus Helper, Prioritizer, Workflow Optimizer

🤖 AI & Tech (8)

Prompt Engineer, Tech Support, Cybersecurity, AI Explainer, Blockchain Expert, Cloud Architect, Network Expert, Database Expert

🏥 Health (5)

Fitness Coach, Nutrition Advisor, Wellness Guide, Sleep Advisor, Meditation Guide

🎮 Fun (8)

Trivia Master, Comedian, Riddler, Game Designer, Debater, Roleplayer, Story Game, Would You Rather

📱 Social Media (6)

Tweet Writer, Instagram Caption, LinkedIn Posts, YouTube Scripts, TikTok Ideas, Social Strategy

🔬 Research (5)

Academic Writer, Literature Review, Hypothesis Generator, Citation Helper, Methodology Advisor

⚖️ Legal (3)

Legal Basics, Privacy Advisor, Terms Writer


🎨 Workflow Templates

DroidClaw includes 6 built-in workflow templates and can create workflows from natural language:

Template Description
☀️ Daily Briefing Every morning: calendar + weather + news summary
📱 Social Media Post Create and post content across platforms
🔬 Research Report Deep research with final synthesized report
📁 File Organizer Organize files by type, date, or content
📧 Email Drafter Draft professional emails from context
📈 Market Monitor Track stocks/crypto and alert on big moves

Create from Natural Language

You: "Every morning at 8am, check my calendar, get the weather, and send me a summary"
DroidClaw: *Creates workflow with 3 steps, schedules it*

🤖 35+ AI Providers

Provider Models Type
🟢 OpenAI GPT-4o, GPT-4o Mini, o1, o3-mini Cloud
🟣 Anthropic Claude Sonnet 4, Claude 3.5 Sonnet/Haiku Cloud
🔵 Google AI Gemini 2.5 Pro/Flash, 2.0 Flash Cloud
🔗 OpenRouter 100+ models (all providers) Cloud
🟢 Nvidia NIM Llama 3.1 405B/70B, DeepSeek R1 Cloud
🟠 Mistral Large, Medium, Small, Codestral Cloud
🔷 DeepSeek V3, Coder, R1 Reasoner Cloud
🟤 Alibaba Qwen Max/Plus/Turbo/VL/Coder Cloud
📱 Xiaomi MiMo V2 Pro, V2 Lite Cloud
🔥 Groq Llama 3.1/3.3 70B, DeepSeek R1 Cloud
🟡 Cohere Command R+, R, A Cloud
🟪 Perplexity Sonar Pro, Sonar Large/Small Cloud
🌐 Together AI Llama 3.1 405B, DeepSeek R1 Cloud
🧠 Cerebras Llama 3.1/3.3 8B/70B Cloud
🟤 SambaNova Llama 3.1 405B, DeepSeek R1 Cloud
🆆 Fireworks Llama 3.1 405B, DeepSeek R1 Cloud
✍️ Writer Palmyra X 004 Cloud
🟦 Replicate Llama 3.1 405B/70B Cloud
🟩 AWS Bedrock Claude 3.5, Llama 3.1 Cloud
🔶 Google Vertex Gemini 2.5 Pro Cloud
🌙 Moonshot Kimi 128K, 32K Cloud
🟫 Zhipu GLM-4 Plus, Flash Cloud
🔶 Baichuan Baichuan 4, 3 Turbo Cloud
🟨 Yi 01.AI Yi Large, Medium, Vision Cloud
🟥 MiniMax Abab 6.5 Cloud

Local Models (Offline)

Family Models Size
🦙 Llama 3.2 1B/3B, 3.1 8B 700MB - 4.7GB
🔵 Gemma 1B, 4B, 12B, 27B 800MB - 16GB
🟦 Phi Phi-4 Mini, Phi-3.5 Mini 2.3-2.4GB
🟤 Qwen 0.5B - 7B 400MB - 4.5GB
🟠 Mistral 7B, Nemo 12B 4.1-7GB
🔷 DeepSeek R1 1.5B, 7B 1.1-4.5GB

🏗️ Architecture

┌─────────────────────────────────────────────────────────────────┐
│                    DroidClaw Gateway (Brain)                     │
│  ┌──────────┐  ┌──────────┐  ┌──────────┐  ┌──────────┐       │
│  │ Provider  │  │  Agent   │  │  Sub-    │  │  Auto-   │       │
│  │ Manager   │──│ Runtime  │──│  Agent   │──│  Creator │       │
│  │ (35+ AI)  │  │ (Loop)   │  │  Manager │  │  Engine  │       │
│  └──────────┘  └──────────┘  └──────────┘  └──────────┘       │
│       │             │             │             │                │
│  ┌──────────┐  ┌──────────┐  ┌──────────┐  ┌──────────┐       │
│  │  Tool    │  │  Skill   │  │ Browser  │  │ Research │       │
│  │  Engine  │  │  Engine  │  │  Engine  │  │  Engine  │       │
│  │ (90+)    │  │ (110+)   │  │ (WebView)│  │ (Deep)   │       │
│  └──────────┘  └──────────┘  └──────────┘  └──────────┘       │
│       │             │             │             │                │
│  ┌──────────┐  ┌──────────┐  ┌──────────┐  ┌──────────┐       │
│  │  Multi-  │  │ Workflow │  │ Heartbeat│  │ Stream-  │       │
│  │  Modal   │  │  Builder │  │  Engine  │  │  ing     │       │
│  │  Engine  │  │ (Visual) │  │ (Proact.)│  │  Engine  │       │
│  └──────────┘  └──────────┘  └──────────┘  └──────────┘       │
│       │             │             │             │                │
│  ┌──────────┐  ┌──────────┐  ┌──────────┐  ┌──────────┐       │
│  │  Voice   │  │  Memory  │  │  File    │  │  Local   │       │
│  │  Engine  │  │  Engine  │  │  Manager │  │  Models  │       │
│  │ (STT/TTS)│  │ (SQLite) │  │          │  │ (20+)    │       │
│  └──────────┘  └──────────┘  └──────────┘  └──────────┘       │
└─────────────────────────────────────────────────────────────────┘

📱 UI Screens

Screen Description
💬 Chat Main agent conversation with tool tracking
📊 Dashboard Live status of agents, tools, memory, heartbeat
Skills Browse, search, enable/disable 110+ skills
🔧 Tools View all 90+ tools by category
🌐 Browser Live WebView with URL bar and console
🧠 Memory Long-term memory entries and knowledge base
🔄 Automation Cron jobs, reminders, IFTTT rules
⚙️ Settings API keys, providers, voice, security, appearance

🛠️ Build From Source

Prerequisites

# Flutter SDK 3.x
brew install flutter  # macOS
snap install flutter --classic  # Linux

# Android Studio
# Download from https://developer.android.com/studio

# Java 17
brew install openjdk@17  # macOS

Build

# Clone
git clone https://github.com/Atum246/droidclaw.git
cd droidclaw

# Dependencies
flutter pub get

# Debug build
flutter run

# Release APK
flutter build apk --release

# Split APKs (smaller per-architecture)
flutter build apk --release --split-per-abi

# Output
ls build/app/outputs/flutter-apk/

GitHub Actions CI/CD

Every push to main automatically builds the APK. Tag a release with v* to:

  • Build universal + split APKs
  • Create GitHub Release with APKs attached
  • Download from Actions tab or Releases tab

⚙️ Configuration

API Keys

  1. Open DroidClaw → Settings ⚙️
  2. Go to API Keys
  3. Enter keys for your providers:
    • OpenAI: sk-...
    • Anthropic: sk-ant-...
    • Google: AIza...
    • OpenRouter: sk-or-...
    • Any of the 35+ supported providers

Local Models (Offline)

  1. Settings → Local Models
  2. Browse available models (Llama, Gemma, Phi, Qwen, etc.)
  3. Download one (e.g., Gemma 3 4B)
  4. Set as active model
  5. Chat without internet!

Remote Control

  1. Enable SSH on your laptop/PC
  2. Use remote_connect tool
  3. Enter host, port, credentials
  4. Control your computer from your phone!

📂 Project Structure

droidclaw/
├── lib/
│   ├── main.dart                              # Entry point
│   ├── core/
│   │   ├── gateway/droidclaw_gateway.dart     # 🧠 Autonomous brain
│   │   ├── providers/ai_provider_manager.dart # 🤖 35+ AI providers
│   │   ├── tools/tool_engine.dart             # 🔧 90+ tools
│   │   ├── tools/tool_creator.dart            # 🔧 Auto-create tools
│   │   ├── skills/skill_engine.dart           # ⚡ 110+ skills
│   │   ├── skills/skill_creator.dart          # ⚡ Auto-create skills
│   │   ├── agents/sub_agent_manager.dart      # 🤖 Sub-agent spawning
│   │   ├── browser/browser_engine.dart        # 🌐 Browser automation
│   │   ├── research/deep_research_engine.dart # 🔬 Deep research
│   │   ├── heartbeat/heartbeat_engine.dart    # 💓 Proactive checks
│   │   ├── streaming/streaming_engine.dart    # 📡 Real-time streaming
│   │   ├── multimodal/multimodal_engine.dart  # 🖼️ Image/voice/doc
│   │   ├── workflow/workflow_builder.dart     # 🎨 Visual workflows
│   │   ├── memory/memory_engine.dart          # 🧠 SQLite memory
│   │   ├── voice/voice_engine.dart            # 🎤 STT & TTS
│   │   ├── automation/automation_engine.dart  # 🔄 Cron & tasks
│   │   ├── search/web_search_engine.dart      # 🌐 Web search
│   │   ├── models/local_model_manager.dart    # 📥 Local models
│   │   └── files/file_manager.dart            # 📁 File management
│   └── ui/
│       ├── theme/droid_theme.dart             # 🎨 Cyberpunk theme
│       └── screens/
│           ├── home_screen.dart               # 🏠 Navigation
│           ├── chat_screen.dart               # 💬 Agent chat
│           ├── dashboard_screen.dart          # 📊 Live dashboard
│           ├── skills_screen.dart             # ⚡ Skills
│           ├── tools_screen.dart              # 🔧 Tools
│           ├── browser_screen.dart            # 🌐 Browser viewer
│           ├── memory_screen.dart             # 🧠 Memory
│           ├── automation_screen.dart         # 🔄 Automation
│           ├── settings_screen.dart           # ⚙️ Settings
│           ├── onboarding_screen.dart         # 🎉 First-run setup
│           └── local_models_screen.dart       # 📥 Downloads
├── .github/workflows/build.yml                # CI/CD
├── android/                                   # Android config
├── assets/                                    # Icons, fonts
├── pubspec.yaml                               # Dependencies
└── README.md                                  # This file

🤝 Contributing

  1. Fork the repository
  2. Create a feature branch: git checkout -b feature/amazing
  3. Commit changes: git commit -m 'Add amazing feature'
  4. Push: git push origin feature/amazing
  5. Open a Pull Request

📄 License

MIT License — Inspired by OpenClaw


🙏 Credits


🤖 DroidClaw — The power of OpenClaw, in your pocket
Can do ANYTHING on your phone

About

🤖 DroidClaw - The Ultimate AI Agent for Android. OpenClaw in your pocket. 35+ AI providers, 80+ tools, 110+ skills, local model downloads, remote control, and more.

Resources

Stars

Watchers

Forks

Packages

 
 
 

Contributors

Languages