RIRD CLI

AI-powered agent for browser automation, code generation, and computer use via command line.

``
_____ _ _______ ______
| _ || | | _ \ | __ \
| |_| || | | |_| | | | \ |
| _ || | | _ < | | | |
| | | || |_____ | |_| | | |__/ /
|_| |_||_______||_____/ |______/

RIRD AI Agent - v2.1.239
`

Features

$3

- AI-Powered Task Execution: Describe what you want, RIRD figures out how to do it
- Browser Automation: Control web browsers programmatically with full anti-detection
- Code Generation: Write, edit, and fix code in any language
- Computer Control: Use mouse, keyboard, take screenshots, manage windows
- File Operations: Read, write, edit files with context-aware changes
- Multi-Model Support: Anthropic, OpenAI, Google, Deepseek, and more

$3

- 12+ Anti-Detection Mechanisms: Bypass bot protection and CAPTCHAs
- Real-time Streaming: See results as they're generated
- Session Management: Save and resume conversations
- Cloud Execution: Run tasks on cloud servers
- Offline Mode: Works without internet (24-hour cache)
- Parallel Execution: Process multiple tasks concurrently
- API Server Mode: REST API for remote task submission
- Performance Tuning: Configure for speed, accuracy, or reliability

Quick Start

$3

macOS/Linux/WSL:
`bash
curl -fsSL https://rird.ai/install.sh | bash
`

Windows (with npm):
`bash
npm install -g rird-ai
`

or:
`bash


Use Git Bash


curl -fsSL https://rird.ai/install.sh | bash
`

$3

Get your license key from rird.ai:

`bash
rird activate YOUR_LICENSE_KEY
`

$3

Interactive mode (recommended):
`bash
rird
`

Then type: Write a Python script that validates email addresses

Single command:
`bash
rird "What is 2 + 2?"
`

$3

Edit ~/.rird/engine/config.yaml to:
- Change LLM model
- Configure API keys
- Adjust browser settings
- Enable cloud execution

Common Use Cases

$3

`bash
rird "Scrape all product names and prices from amazon.com and save to CSV"
`

$3

`bash
rird "Generate a React component for user authentication with email and password fields"
`

$3

`bash
rird "Extract all contact information from my PDF resume and format as JSON"
`

$3

`bash
rird "Test my website login flow, including forgotten password recovery"
`

$3

`bash
rird "Analyze this CSV file and create a summary report with key metrics"
`

$3

`bash
rird "Find all jpg files in ~/Downloads, rename them with timestamps, and organize into folders by month"
`

Command Reference

For detailed command reference, see docs/CLI_REFERENCE.md

`bash
rird # Start interactive mode
rird "your task here" # Run single task
rird help # Show all commands
rird --version # Show version
rird activate # Activate license
rird status # Check license status
rird models # List available models
rird health-check # Verify installation
rird debug # Show debug info
rird serve --port 9000 # Start API server
`

Installation & Setup

$3

- RAM: 2GB minimum, 4GB+ recommended
- Disk: 500MB for RIRD + 1GB for cache
- Network: Internet connection for LLM API
- Python: 3.8+ (installed automatically)
- Platforms: macOS, Linux, Windows, WSL

For detailed setup guide, see docs/PRODUCTION_SETUP.md

Configuration

Edit ~/.rird/engine/config.yaml to customize:

`yaml
llm:
mode: remote
remote_url: https://rird.ai/api/llm
main_model: 0000/ui-tars-1.5-7b:latest

browser:
headless: true # Show browser? (true/false)
timeout: 30000 # Request timeout (ms)
stealth_mode: true # Anti-detection

execution:
max_retries: 3
task_timeout: 300000 # 5 minutes

logging:
level: info # debug, info, warn, error
file: ~/.rird/logs/rird.log
`

API Keys

Set your LLM provider keys via environment variables:

`bash
export ANTHROPIC_API_KEY="sk-ant-..."
export OPENAI_API_KEY="sk-..."
export GOOGLE_GENERATIVE_AI_API_KEY="..."
rird "your task"
`

Or edit ~/.rird/engine/rird.json for persistent configuration.

Global Flags

These flags work with any command:

`bash
rird --help # Show command help
rird -q # Quiet mode (less output)
rird --json # Output as JSON
rird --no-color # Disable colored output
DEBUG=1 rird # Enable debug logging
`

Advanced Usage

$3

Execute tasks on cloud servers:

`bash
rird --cloud "Complex analysis that needs more resources"
`

$3

Run as HTTP API server:

`bash
rird serve --port 8765
`

Then submit tasks:
`bash
curl -X POST http://localhost:8765/api/task \
-H "Authorization: Bearer YOUR_TOKEN" \
-d '{"task": "What is 2+2?"}'
`

$3

Process multiple tasks:

`bash
while read task; do
rird "$task" --json
done < tasks.txt
`

$3

`yaml
name: RIRD Automation
on: [push]
jobs:
task:
runs-on: ubuntu-latest
steps:
- run: curl -fsSL https://rird.ai/install.sh | bash
- run: rird activate ${{ secrets.RIRD_KEY }}
- run: rird "Run automated tests" --json
env:
OPENAI_API_KEY: ${{ secrets.OPENAI_KEY }}
`

Troubleshooting

$3

License validation fails:
`bash


Clear cache and re-validate


rm ~/.rird/license_cache.json
rird status
`

Out of memory:
`bash
NODE_OPTIONS=--max-old-space-size=4096 rird
`

Network error:
`bash


Check connectivity

ping rird.ai

Use proxy if needed

export RIRD_PROXY="http://proxy:8080"
`

For more help, see docs/TROUBLESHOOTING.md or run:

`bash
rird health-check
`

Documentation

- PRODUCTION_SETUP.md - Installation, configuration, scaling
- CLI_REFERENCE.md - Complete command reference
- ARCHITECTURE.md - System design and internals
- TROUBLESHOOTING.md - Problem-solving guide

System Architecture

RIRD consists of:

1. CLI/TUI - Terminal user interface for interactive mode
2. Agent Engine - AI planning and task execution
3. Tool Execution - Browser automation, file I/O, computer control
4. LLM Integration - Support for multiple AI providers
5. License System - Device-bound license validation
6. Configuration - YAML-based settings management

See docs/ARCHITECTURE.md for detailed technical design.

Security

$3

- Local-first: Conversations saved locally only
- Device-bound licenses: Prevent key sharing
- No telemetry: Unless explicitly enabled
- Encrypted cache: 24-hour offline capability

$3

RIRD implements 12+ techniques to bypass bot detection:
- Fingerprint spoofing
- Human-like timing and mouse movement
- User-agent rotation
- Proxy support
- WebDriver detection evasion
- Headless detection mitigation

See docs/ARCHITECTURE.md Security section for details.

Update and Support

$3

RIRD automatically checks for updates (can be disabled):

`bash
rird upgrade # Manual update
export RIRD_SKIP_AUTO_UPDATE=1 # Disable auto-updates
`

$3

- Documentation: https://rird.ai/docs
- GitHub Issues: https://github.com/rird-ai/rird/issues
- Community Forum: https://community.rird.ai
- Email Support: support@rird.ai
- Status Page: https://status.rird.ai

$3

Run diagnostic command and include output:

`bash
rird health-check
DEBUG=1 rird 2>&1 | head -200
`

Roadmap

- [ ] Plugin system for custom tools
- [ ] Distributed execution across multiple machines
- [ ] Real-time multi-user collaboration
- [ ] Advanced prompt caching
- [ ] Custom model fine-tuning
- [ ] Multi-agent coordination

License

RIRD requires a valid license key for production use.

- Free Trial: 7 days, limited features
- Pro: $49/month, full access
- Enterprise: Custom pricing

Get started: https://rird.ai

Contributing

RIRD is built on open-source foundations (OpenCode fork). To contribute:

1. Fork: https://github.com/rird-ai/rird
2. Create feature branch: git checkout -b feature/amazing
3. Commit changes: git commit -am 'Add feature'
4. Push: git push origin feature/amazing
5. Open pull request

Credits

RIRD is forked from OpenCode, an open-source AI coding agent.

$3

- Bun - Fast JavaScript runtime and package manager
- TypeScript - Type-safe development
- Claude AI - Primary LLM backend
- Playwright - Browser automation
- MCP Protocol - Tool integration standard
- Drizzle ORM - Database access

---

Ready to get started?

`bash
curl -fsSL https://rird.ai/install.sh | bash
rird activate YOUR_LICENSE_KEY
rird
``

Questions? Email support@rird.ai or visit https://rird.ai/docs