CD/CD consolidation

etc/llm-providers/claude.yaml

@@ -0,0 +1,84 @@
+# Claude (Anthropic) LLM Provider Configuration
+# Documentation: https://docs.anthropic.com/en/api
+
+# Provider metadata
+provider:
+  id: claude
+  name: Claude
+  description: Anthropic Claude models via API
+
+# Enable/disable this provider
+enabled: true
+
+# Priority for provider selection (lower = higher priority)
+priority: 100
+
+# API Configuration
+api:
+  # API key (can also use ANTHROPIC_API_KEY environment variable)
+  apiKey: "${ANTHROPIC_API_KEY}"
+
+  # Base URL for API requests
+  baseUrl: "https://api.anthropic.com"
+
+  # API version header
+  apiVersion: "2023-06-01"
+
+# Model Configuration
+model:
+  # Model to use for inference
+  # Options: claude-sonnet-4-20250514, claude-opus-4-20250514, claude-3-5-haiku-20241022
+  name: "claude-sonnet-4-20250514"
+
+  # Fallback models if primary is unavailable
+  fallbacks:
+    - "claude-3-5-haiku-20241022"
+
+# Inference Parameters
+inference:
+  # Temperature (0 = deterministic, 1 = creative)
+  temperature: 0
+
+  # Maximum tokens to generate
+  maxTokens: 4096
+
+  # Top-p (nucleus sampling)
+  topP: 1.0
+
+  # Top-k sampling (0 = disabled)
+  topK: 0
+
+  # Stop sequences
+  stopSequences: []
+
+# Request Configuration
+request:
+  # Request timeout
+  timeout: "00:02:00"
+
+  # Maximum retries on failure
+  maxRetries: 3
+
+  # Retry delay
+  retryDelay: "00:00:01"
+
+# Rate Limiting
+rateLimit:
+  # Requests per minute (0 = unlimited)
+  requestsPerMinute: 0
+
+  # Tokens per minute (0 = unlimited)
+  tokensPerMinute: 0
+
+# Extended Thinking (Claude 3.5+ feature)
+thinking:
+  # Enable extended thinking for complex reasoning
+  enabled: false
+
+  # Budget tokens for thinking (when enabled)
+  budgetTokens: 10000
+
+# Logging
+logging:
+  logBodies: false
+  logUsage: true