aidp 0.7.0 → 0.8.0

data/templates/aidp-production.yml.example

@@ -0,0 +1,475 @@
+# Production AIDP Configuration
+# This is a production-ready configuration with comprehensive monitoring and security
+# Copy this file to your project root as 'aidp.yml' and customize as needed
+
+# Harness configuration
+harness:
+  # Default provider to use
+  default_provider: "claude"
+
+  # Fallback providers in order of preference
+  fallback_providers: ["cursor", "gemini"]
+
+  # Production settings
+  max_retries: 3
+  request_timeout: 300
+  user_feedback_timeout: 300
+  work_completion_timeout: 1800
+
+  # Advanced features
+  auto_switch_on_error: true
+  auto_switch_on_rate_limit: true
+  max_concurrent_requests: 10
+
+  # Circuit breaker configuration
+  circuit_breaker:
+    enabled: true
+    failure_threshold: 5
+    timeout: 300
+    half_open_max_calls: 3
+
+  # Retry configuration
+  retry:
+    enabled: true
+    max_attempts: 3
+    base_delay: 2.0
+    max_delay: 120.0
+    exponential_base: 2.0
+    jitter: true
+
+  # Rate limiting configuration
+  rate_limit:
+    enabled: true
+    default_reset_time: 3600
+    burst_limit: 20
+    sustained_limit: 10
+
+  # Load balancing configuration
+  load_balancing:
+    enabled: true
+    strategy: "weighted_round_robin"
+    health_check_interval: 30
+    unhealthy_threshold: 3
+
+  # Health check configuration
+  health_check:
+    enabled: true
+    interval: 60
+    timeout: 15
+    failure_threshold: 3
+    success_threshold: 2
+
+  # Metrics configuration
+  metrics:
+    enabled: true
+    retention_days: 90
+    aggregation_interval: 300
+    export_interval: 3600
+
+  # Session configuration
+  session:
+    enabled: true
+    timeout: 3600
+    sticky_sessions: true
+    session_affinity: "provider_model"
+
+  # User interface configuration
+  ui:
+    enabled: true
+    interactive_mode: true
+    file_selection: true
+    progress_display: true
+    status_updates: true
+    pause_resume: true
+
+  # Error handling configuration
+  error_handling:
+    enabled: true
+    log_errors: true
+    retry_on_error: true
+    fallback_on_error: true
+    error_notification: true
+
+  # Performance optimization
+  performance:
+    enabled: true
+    cache_responses: true
+    parallel_processing: true
+    batch_requests: true
+    connection_pooling: true
+
+  # Security configuration
+  security:
+    enabled: true
+    ssl_verify: true
+    allowed_hosts: []
+    blocked_hosts: []
+    api_key_rotation: true
+    audit_logging: true
+
+# Provider configurations
+providers:
+  # Claude provider (primary)
+  claude:
+    type: "api"
+    priority: 1
+    max_tokens: 200000
+    default_flags: ["--dangerously-skip-permissions"]
+
+    # Available models
+    models: ["claude-3-5-sonnet-20241022", "claude-3-5-haiku-20241022"]
+
+    # Model weights for load balancing
+    model_weights:
+      claude-3-5-sonnet-20241022: 3
+      claude-3-5-haiku-20241022: 2
+
+    # Model-specific configurations
+    models_config:
+      claude-3-5-sonnet-20241022:
+        flags: ["--dangerously-skip-permissions"]
+        max_tokens: 200000
+        timeout: 300
+      claude-3-5-haiku-20241022:
+        flags: ["--dangerously-skip-permissions"]
+        max_tokens: 200000
+        timeout: 180
+
+    # Authentication configuration
+    auth:
+      api_key_env: "ANTHROPIC_API_KEY"
+
+    # Endpoint configuration
+    endpoints:
+      default: "https://api.anthropic.com/v1/messages"
+
+    # Provider features
+    features:
+      file_upload: true
+      code_generation: true
+      analysis: true
+      vision: true
+      streaming: true
+      function_calling: true
+      tool_use: true
+
+    # Monitoring configuration
+    monitoring:
+      enabled: true
+      metrics_interval: 30
+
+    # Rate limiting configuration
+    rate_limit:
+      enabled: true
+      requests_per_minute: 50
+      requests_per_hour: 5000
+      tokens_per_minute: 40000
+      tokens_per_hour: 2000000
+      burst_limit: 5
+
+    # Retry configuration
+    retry:
+      enabled: true
+      max_attempts: 3
+      base_delay: 2.0
+      max_delay: 120.0
+      exponential_base: 2.0
+      jitter: true
+      retry_on_rate_limit: true
+
+    # Circuit breaker configuration
+    circuit_breaker:
+      enabled: true
+      failure_threshold: 5
+      timeout: 300
+      half_open_max_calls: 3
+
+    # Cost tracking
+    cost:
+      input_cost_per_token: 0.000003
+      output_cost_per_token: 0.000015
+      fixed_cost_per_request: 0.0
+      currency: "USD"
+
+    # Health check configuration
+    health_check:
+      enabled: true
+      interval: 60
+      timeout: 15
+      failure_threshold: 3
+      success_threshold: 2
+      check_url: "https://api.anthropic.com/v1/messages"
+
+    # Logging configuration
+    log:
+      enabled: true
+      level: "info"
+      file: "logs/claude.log"
+      max_size: 52428800  # 50MB
+      max_files: 10
+      format: "json"
+      log_requests: true
+      log_responses: false
+
+    # Cache configuration
+    cache:
+      enabled: true
+      ttl: 1800
+      max_size: 500
+      strategy: "lru"
+
+    # Security configuration
+    security:
+      ssl_verify: true
+      allowed_hosts: ["api.anthropic.com"]
+      blocked_hosts: []
+      timeout: 30
+      max_redirects: 5
+
+  # Cursor provider (fallback)
+  cursor:
+    type: "package"
+    priority: 2
+    default_flags: []
+
+    # Available models
+    models: ["cursor-default", "cursor-fast"]
+
+    # Model weights for load balancing
+    model_weights:
+      cursor-default: 3
+      cursor-fast: 2
+
+    # Model-specific configurations
+    models_config:
+      cursor-default:
+        flags: []
+        timeout: 600
+      cursor-fast:
+        flags: ["--fast"]
+        timeout: 300
+
+    # Provider features
+    features:
+      file_upload: true
+      code_generation: true
+      analysis: true
+      streaming: true
+
+    # Monitoring configuration
+    monitoring:
+      enabled: true
+      metrics_interval: 30
+
+    # Rate limiting configuration
+    rate_limit:
+      enabled: true
+      requests_per_minute: 60
+      requests_per_hour: 1000
+      burst_limit: 10
+
+    # Retry configuration
+    retry:
+      enabled: true
+      max_attempts: 3
+      base_delay: 1.0
+      max_delay: 60.0
+      exponential_base: 2.0
+      jitter: true
+
+    # Circuit breaker configuration
+    circuit_breaker:
+      enabled: true
+      failure_threshold: 5
+      timeout: 300
+      half_open_max_calls: 3
+
+    # Cost tracking
+    cost:
+      input_cost_per_token: 0.0
+      output_cost_per_token: 0.0
+      fixed_cost_per_request: 0.0
+      currency: "USD"
+
+    # Health check configuration
+    health_check:
+      enabled: true
+      interval: 60
+      timeout: 10
+      failure_threshold: 3
+      success_threshold: 2
+
+    # Logging configuration
+    log:
+      enabled: true
+      level: "info"
+      file: "logs/cursor.log"
+      max_size: 52428800  # 50MB
+      max_files: 10
+      format: "json"
+      log_requests: true
+      log_responses: false
+
+    # Cache configuration
+    cache:
+      enabled: true
+      ttl: 3600
+      max_size: 200
+      strategy: "lru"
+
+    # Security configuration
+    security:
+      ssl_verify: true
+      allowed_hosts: []
+      blocked_hosts: []
+      timeout: 30
+      max_redirects: 5
+
+  # Gemini provider (fallback)
+  gemini:
+    type: "api"
+    priority: 3
+    max_tokens: 100000
+    default_flags: []
+
+    # Available models
+    models: ["gemini-1.5-pro", "gemini-1.5-flash"]
+
+    # Model weights for load balancing
+    model_weights:
+      gemini-1.5-pro: 3
+      gemini-1.5-flash: 2
+
+    # Model-specific configurations
+    models_config:
+      gemini-1.5-pro:
+        flags: []
+        max_tokens: 100000
+        timeout: 300
+      gemini-1.5-flash:
+        flags: []
+        max_tokens: 100000
+        timeout: 180
+
+    # Authentication configuration
+    auth:
+      api_key_env: "GEMINI_API_KEY"
+
+    # Endpoint configuration
+    endpoints:
+      default: "https://generativelanguage.googleapis.com/v1beta/models"
+
+    # Provider features
+    features:
+      file_upload: true
+      code_generation: true
+      analysis: true
+      vision: true
+      streaming: true
+
+    # Monitoring configuration
+    monitoring:
+      enabled: true
+      metrics_interval: 30
+
+    # Rate limiting configuration
+    rate_limit:
+      enabled: true
+      requests_per_minute: 60
+      requests_per_hour: 1500
+      tokens_per_minute: 32000
+      tokens_per_hour: 1000000
+      burst_limit: 10
+
+    # Retry configuration
+    retry:
+      enabled: true
+      max_attempts: 3
+      base_delay: 1.0
+      max_delay: 60.0
+      exponential_base: 2.0
+      jitter: true
+      retry_on_rate_limit: true
+
+    # Circuit breaker configuration
+    circuit_breaker:
+      enabled: true
+      failure_threshold: 5
+      timeout: 300
+      half_open_max_calls: 3
+
+    # Cost tracking
+    cost:
+      input_cost_per_token: 0.00000125
+      output_cost_per_token: 0.000005
+      fixed_cost_per_request: 0.0
+      currency: "USD"
+
+    # Health check configuration
+    health_check:
+      enabled: true
+      interval: 60
+      timeout: 15
+      failure_threshold: 3
+      success_threshold: 2
+      check_url: "https://generativelanguage.googleapis.com/v1beta/models"
+
+    # Logging configuration
+    log:
+      enabled: true
+      level: "info"
+      file: "logs/gemini.log"
+      max_size: 52428800  # 50MB
+      max_files: 10
+      format: "json"
+      log_requests: true
+      log_responses: false
+
+    # Cache configuration
+    cache:
+      enabled: true
+      ttl: 1800
+      max_size: 500
+      strategy: "lru"
+
+    # Security configuration
+    security:
+      ssl_verify: true
+      allowed_hosts: ["generativelanguage.googleapis.com"]
+      blocked_hosts: []
+      timeout: 30
+      max_redirects: 5
+
+# Environment-specific configurations
+environments:
+  production:
+    harness:
+      max_retries: 3
+      request_timeout: 300
+    providers:
+      claude:
+        rate_limit:
+          requests_per_minute: 50
+        log:
+          level: "info"
+      cursor:
+        log:
+          level: "warn"
+      gemini:
+        rate_limit:
+          requests_per_minute: 60
+        log:
+          level: "info"
+
+# Configuration tips for production:
+# - Set up proper API keys in environment variables
+# - Configure monitoring and alerting
+# - Set up log rotation and retention
+# - Configure rate limits based on your API plans
+# - Enable circuit breakers for fault tolerance
+# - Use load balancing for high availability
+# - Configure health checks for all providers
+# - Set up cost tracking and budgeting
+# - Enable audit logging for compliance
+# - Configure security settings appropriately
+# - Test failover scenarios regularly