Skip to content

03-configuration.md

Latest commit

 

History

History
294 lines (216 loc) · 7.38 KB

03-configuration.md

File metadata and controls

294 lines (216 loc) · 7.38 KB
Raw

User Configuration Guide

Simple, user-friendly configuration options for AutoBot. For detailed technical configuration, see Complete Configuration Reference.

Quick Configuration

Accessing Settings

AutoBot provides multiple ways to configure settings:

  1. Web Interface: Click the gear icon (⚙️) in the web interface
  2. Chat Commands: Type /settings in the chat
  3. Configuration File: Edit config/config.yaml directly

Essential Settings

Choose Your AI Provider

Option 1: Ollama (Local, Free)

llm_config:
  default_llm: "ollama_tinyllama"
  ollama:
    host: "http://localhost:11434"

Option 2: OpenAI (Cloud, Paid)

llm_config:
  default_llm: "openai_gpt35"
  openai:
    api_key: ""  # Set via environment variable OPENAI_API_KEY

Option 3: Anthropic Claude (Cloud, Paid)

llm_config:
  default_llm: "anthropic_claude"
  anthropic:
    api_key: ""  # Set via environment variable ANTHROPIC_API_KEY

Basic System Settings

# Change ports if needed
backend:
  server_port: 8001

# Frontend port (if running development server)
frontend:
  port: 5173

# Enable/disable features
memory:
  redis:
    enabled: true    # Set false if Redis not available
  chromadb:
    enabled: true    # Vector database for knowledge search

Configuration Through Web Interface

LLM Settings

  1. Access Settings: Click ⚙️ icon → "LLM Configuration"
  2. Choose Provider: Select Ollama, OpenAI, or Anthropic
  3. Select Model: Pick from available models
  4. Adjust Temperature:
    • Low (0.1-0.3): More focused, deterministic responses
    • Medium (0.4-0.7): Balanced creativity and consistency
    • High (0.8-1.0): More creative, varied responses

System Behavior

  1. Access Settings: Click ⚙️ icon → "Agent Behavior"
  2. Autonomy Level:
    • Supervised: Asks permission for system commands (recommended)
    • Autonomous: Executes approved commands automatically
    • Restricted: Chat only, no system access
  3. Command Permissions: Enable/disable file system, network, and command execution

Performance Settings

  1. Access Settings: Click ⚙️ icon → "Performance"
  2. Hardware Acceleration: Enable GPU if available
  3. Memory Limits: Set RAM usage limits
  4. Model Options: Choose faster vs higher quality models

Common Configuration Scenarios

Home User Setup (Recommended)

llm_config:
  default_llm: "ollama_tinyllama"  # Fast, local, free

agent:
  autonomy_level: "supervised"     # Safe, asks permission

memory:
  redis:
    enabled: true                  # Enhanced features
  chromadb:
    enabled: true                  # Knowledge search

hardware:
  gpu:
    enabled: false                 # Use CPU (stable)

Developer Setup

llm_config:
  default_llm: "ollama_llama2"     # Better code understanding

agent:
  autonomy_level: "autonomous"     # Faster workflow
  max_task_depth: 15               # Complex tasks

debugging:
  enabled: true                    # Development tools
  log_level: "debug"               # Detailed logs

hardware:
  gpu:
    enabled: true                  # Faster inference

Enterprise Setup

llm_config:
  default_llm: "openai_gpt4"       # Highest quality

security:
  strict_mode: true                # Enhanced security
  command_filter:
    enabled: true                  # Filter dangerous commands

logging:
  audit_enabled: true              # Compliance logging
  retention_days: 90               # Keep logs longer

network:
  require_auth: true               # Authentication required

Environment Variables

For sensitive information, use environment variables instead of config files:

# AI Provider API Keys
export OPENAI_API_KEY="your-openai-key"  # pragma: allowlist secret
export ANTHROPIC_API_KEY="your-anthropic-key"  # pragma: allowlist secret

# Server Configuration
export AUTOBOT_BACKEND_PORT=8001
export AUTOBOT_FRONTEND_PORT=5173

# Redis Configuration
export AUTOBOT_REDIS_HOST="localhost"
export AUTOBOT_REDIS_PORT=6379
export AUTOBOT_REDIS_PASSWORD="your-password"  # pragma: allowlist secret

# Security
export AUTOBOT_SECRET_KEY="your-secure-secret"  # pragma: allowlist secret

Add these to your shell profile (.bashrc, .zshrc) to persist across sessions.

Configuration Validation

Check Configuration

# Validate configuration file
python -c "from src.config import config; print('✓ Configuration valid')"

# Test specific components
python -c "from src.diagnostics import check_llm; check_llm()"
python -c "from src.diagnostics import check_redis; check_redis()"

Configuration Status

Use the web interface or chat commands:

/status config

/settings validate

Backup and Restore

Backup Current Configuration

# Create timestamped backup
cp config/config.yaml config/config.backup.$(date +%Y%m%d_%H%M%S).yaml

# Quick backup
cp config/config.yaml config/config.backup.yaml

Restore Configuration

# Restore from backup
cp config/config.backup.yaml config/config.yaml

# Reset to defaults
cp config/config.yaml.template config/config.yaml

Export Settings

Through web interface:

  1. Settings → "Export Configuration"
  2. Save the downloaded file as backup
  3. Import on other AutoBot instances

Troubleshooting Configuration

Common Issues

Issue: Changes not taking effect

  • Restart AutoBot: ./run_agent.sh
  • Check configuration syntax: python -c "import yaml; yaml.safe_load(open('config/config.yaml'))"

Issue: LLM connection failed

  • Verify API keys in environment variables
  • For Ollama: ensure service is running (ollama serve)
  • Check network connectivity

Issue: Port conflicts

  • Change ports in configuration:
backend:
  server_port: 8002  # Change from default 8001

Issue: Performance problems

  • Disable GPU if causing issues:
hardware:
  gpu:
    enabled: false
  • Use lighter models (tinyllama instead of llama2)

Configuration Help

  1. Built-in Validation: AutoBot validates configuration on startup
  2. Web Interface: Visual feedback for configuration errors
  3. Log Files: Check logs/autobot.log for configuration issues
  4. Chat Support: Ask AutoBot about configuration: "Help me configure my settings"

Advanced Configuration

For advanced users who need detailed technical configuration:

Configuration Best Practices

Security

  1. Never commit API keys to version control
  2. Use environment variables for sensitive data
  3. Set secure file permissions: chmod 600 config/config.yaml
  4. Regular backups of working configurations

Performance

  1. Start simple: Use default settings first
  2. Monitor resources: Watch CPU/RAM usage
  3. GPU acceleration: Enable only if stable
  4. Model selection: Balance speed vs quality

Maintenance

  1. Version control: Track configuration changes
  2. Documentation: Comment complex settings
  3. Testing: Validate after changes
  4. Monitoring: Watch logs for issues

Need more advanced configuration options? See the Complete Configuration Reference for technical details.