elsabro 2.3.0 → 3.7.0

package/templates/multi-llm-config.json

@@ -0,0 +1,544 @@
+{
+  "multiLLMProviders": {
+    "enabled": true,
+    "version": "1.0.0",
+
+    "defaultModel": "claude-sonnet-4",
+    "defaultProvider": "claude",
+
+    "providers": {
+      "claude": {
+        "enabled": true,
+        "priority": 1,
+        "config": {
+          "apiKey": "${ANTHROPIC_API_KEY}",
+          "baseUrl": "https://api.anthropic.com",
+          "defaultModel": "claude-sonnet-4",
+          "timeout": 120000,
+          "maxRetries": 3,
+          "retryDelay": 1000,
+          "betaHeaders": []
+        },
+        "models": {
+          "claude-opus-4-5": {
+            "id": "claude-opus-4-5-20251101",
+            "contextWindow": 200000,
+            "maxOutput": 32000,
+            "inputPrice": 15.0,
+            "outputPrice": 75.0,
+            "capabilities": ["vision", "tools", "extended_thinking", "streaming"]
+          },
+          "claude-sonnet-4": {
+            "id": "claude-sonnet-4-20250514",
+            "contextWindow": 200000,
+            "maxOutput": 64000,
+            "inputPrice": 3.0,
+            "outputPrice": 15.0,
+            "capabilities": ["vision", "tools", "computer_use", "streaming"]
+          },
+          "claude-3-5-haiku": {
+            "id": "claude-3-5-haiku-20241022",
+            "contextWindow": 200000,
+            "maxOutput": 8192,
+            "inputPrice": 0.80,
+            "outputPrice": 4.0,
+            "capabilities": ["vision", "tools", "streaming"]
+          }
+        },
+        "rateLimit": {
+          "requestsPerMinute": 50,
+          "tokensPerMinute": 100000,
+          "tokensPerDay": 5000000,
+          "concurrent": 5
+        }
+      },
+
+      "openai": {
+        "enabled": true,
+        "priority": 2,
+        "config": {
+          "apiKey": "${OPENAI_API_KEY}",
+          "organization": "${OPENAI_ORG_ID}",
+          "baseUrl": "https://api.openai.com/v1",
+          "defaultModel": "gpt-4o",
+          "timeout": 120000,
+          "maxRetries": 3,
+          "retryDelay": 1000
+        },
+        "models": {
+          "gpt-4o": {
+            "id": "gpt-4o",
+            "contextWindow": 128000,
+            "maxOutput": 16384,
+            "inputPrice": 2.50,
+            "outputPrice": 10.0,
+            "capabilities": ["vision", "tools", "json_mode", "streaming"]
+          },
+          "gpt-4-turbo": {
+            "id": "gpt-4-turbo",
+            "contextWindow": 128000,
+            "maxOutput": 4096,
+            "inputPrice": 10.0,
+            "outputPrice": 30.0,
+            "capabilities": ["vision", "tools", "json_mode", "streaming"]
+          },
+          "gpt-4o-mini": {
+            "id": "gpt-4o-mini",
+            "contextWindow": 128000,
+            "maxOutput": 16384,
+            "inputPrice": 0.15,
+            "outputPrice": 0.60,
+            "capabilities": ["vision", "tools", "json_mode", "streaming"]
+          },
+          "gpt-3.5-turbo": {
+            "id": "gpt-3.5-turbo",
+            "contextWindow": 16385,
+            "maxOutput": 4096,
+            "inputPrice": 0.50,
+            "outputPrice": 1.50,
+            "capabilities": ["tools", "json_mode", "streaming"]
+          }
+        },
+        "rateLimit": {
+          "requestsPerMinute": 60,
+          "tokensPerMinute": 150000,
+          "tokensPerDay": 10000000,
+          "concurrent": 10
+        }
+      },
+
+      "gemini": {
+        "enabled": true,
+        "priority": 3,
+        "config": {
+          "apiKey": "${GOOGLE_API_KEY}",
+          "baseUrl": "https://generativelanguage.googleapis.com",
+          "defaultModel": "gemini-2.0-flash",
+          "timeout": 120000,
+          "maxRetries": 3,
+          "retryDelay": 1000
+        },
+        "models": {
+          "gemini-2.0-flash": {
+            "id": "gemini-2.0-flash",
+            "contextWindow": 1000000,
+            "maxOutput": 8192,
+            "inputPrice": 0.075,
+            "outputPrice": 0.30,
+            "capabilities": ["vision", "tools", "grounding", "streaming"]
+          },
+          "gemini-1.5-pro": {
+            "id": "gemini-1.5-pro",
+            "contextWindow": 2000000,
+            "maxOutput": 8192,
+            "inputPrice": 1.25,
+            "outputPrice": 5.0,
+            "capabilities": ["vision", "tools", "grounding", "code_execution", "streaming"]
+          },
+          "gemini-1.5-flash": {
+            "id": "gemini-1.5-flash",
+            "contextWindow": 1000000,
+            "maxOutput": 8192,
+            "inputPrice": 0.075,
+            "outputPrice": 0.30,
+            "capabilities": ["vision", "tools", "streaming"]
+          }
+        },
+        "rateLimit": {
+          "requestsPerMinute": 60,
+          "tokensPerMinute": 1000000,
+          "tokensPerDay": null,
+          "concurrent": 10
+        }
+      },
+
+      "local": {
+        "enabled": false,
+        "priority": 10,
+        "config": {
+          "baseUrl": "http://localhost:11434",
+          "defaultModel": "llama3",
+          "timeout": 300000,
+          "maxRetries": 1,
+          "retryDelay": 0
+        },
+        "models": {
+          "llama3": {
+            "id": "llama3:latest",
+            "contextWindow": 8192,
+            "maxOutput": 4096,
+            "inputPrice": 0,
+            "outputPrice": 0,
+            "capabilities": ["tools"]
+          },
+          "llama3.2": {
+            "id": "llama3.2:latest",
+            "contextWindow": 128000,
+            "maxOutput": 4096,
+            "inputPrice": 0,
+            "outputPrice": 0,
+            "capabilities": ["tools", "vision"]
+          },
+          "mistral": {
+            "id": "mistral:latest",
+            "contextWindow": 32768,
+            "maxOutput": 4096,
+            "inputPrice": 0,
+            "outputPrice": 0,
+            "capabilities": ["tools"]
+          },
+          "codellama": {
+            "id": "codellama:latest",
+            "contextWindow": 16384,
+            "maxOutput": 4096,
+            "inputPrice": 0,
+            "outputPrice": 0,
+            "capabilities": ["code"]
+          },
+          "deepseek-coder": {
+            "id": "deepseek-coder:latest",
+            "contextWindow": 16384,
+            "maxOutput": 4096,
+            "inputPrice": 0,
+            "outputPrice": 0,
+            "capabilities": ["code"]
+          }
+        },
+        "rateLimit": {
+          "requestsPerMinute": null,
+          "tokensPerMinute": null,
+          "tokensPerDay": null,
+          "concurrent": 1
+        }
+      },
+
+      "azure": {
+        "enabled": false,
+        "priority": 2,
+        "config": {
+          "apiKey": "${AZURE_OPENAI_API_KEY}",
+          "endpoint": "${AZURE_OPENAI_ENDPOINT}",
+          "apiVersion": "2024-02-01",
+          "timeout": 120000,
+          "maxRetries": 3,
+          "retryDelay": 1000,
+          "deployments": {
+            "gpt-4o": "gpt-4o-deployment",
+            "gpt-4-turbo": "gpt-4-turbo-deployment",
+            "gpt-3.5-turbo": "gpt-35-turbo-deployment"
+          }
+        },
+        "models": {
+          "gpt-4o": {
+            "contextWindow": 128000,
+            "maxOutput": 16384,
+            "inputPrice": 2.50,
+            "outputPrice": 10.0,
+            "capabilities": ["vision", "tools", "json_mode", "streaming"]
+          },
+          "gpt-4-turbo": {
+            "contextWindow": 128000,
+            "maxOutput": 4096,
+            "inputPrice": 10.0,
+            "outputPrice": 30.0,
+            "capabilities": ["vision", "tools", "json_mode", "streaming"]
+          },
+          "gpt-3.5-turbo": {
+            "contextWindow": 16385,
+            "maxOutput": 4096,
+            "inputPrice": 0.50,
+            "outputPrice": 1.50,
+            "capabilities": ["tools", "json_mode", "streaming"]
+          }
+        },
+        "rateLimit": {
+          "requestsPerMinute": 60,
+          "tokensPerMinute": 150000,
+          "tokensPerDay": null,
+          "concurrent": 10
+        }
+      }
+    },
+
+    "router": {
+      "enabled": true,
+      "strategy": "intelligent",
+
+      "rules": [
+        {
+          "id": "quick-tasks",
+          "name": "Route quick tasks to fast models",
+          "conditions": [
+            { "type": "task_type", "operator": "eq", "value": "quick_task" }
+          ],
+          "target": { "type": "alias", "value": "fast" },
+          "priority": 100,
+          "enabled": true
+        },
+        {
+          "id": "code-generation-large",
+          "name": "Route large code generation to smart models",
+          "conditions": [
+            { "type": "task_type", "operator": "eq", "value": "code_generation" },
+            { "type": "token_count", "operator": "gt", "value": 1000 }
+          ],
+          "target": { "type": "alias", "value": "smart" },
+          "priority": 90,
+          "enabled": true
+        },
+        {
+          "id": "complex-reasoning",
+          "name": "Route complex reasoning to best models",
+          "conditions": [
+            { "type": "task_type", "operator": "eq", "value": "reasoning" }
+          ],
+          "target": { "type": "model", "value": "claude-opus-4-5" },
+          "priority": 95,
+          "enabled": true
+        },
+        {
+          "id": "cost-sensitive",
+          "name": "Route cost-sensitive requests to cheap models",
+          "conditions": [
+            { "type": "cost_limit", "operator": "lt", "value": 0.01 }
+          ],
+          "target": { "type": "alias", "value": "cheap" },
+          "priority": 80,
+          "enabled": true
+        },
+        {
+          "id": "vision-tasks",
+          "name": "Route vision tasks to vision-capable models",
+          "conditions": [
+            { "type": "capability", "operator": "contains", "value": "vision" }
+          ],
+          "target": { "type": "fallback_chain", "value": ["gpt-4o", "claude-sonnet-4", "gemini-2.0-flash"] },
+          "priority": 85,
+          "enabled": true
+        },
+        {
+          "id": "long-context",
+          "name": "Route long context to high-context models",
+          "conditions": [
+            { "type": "token_count", "operator": "gt", "value": 100000 }
+          ],
+          "target": { "type": "fallback_chain", "value": ["gemini-1.5-pro", "claude-sonnet-4"] },
+          "priority": 88,
+          "enabled": true
+        }
+      ],
+
+      "aliases": {
+        "fast": "claude-3-5-haiku",
+        "smart": "claude-sonnet-4",
+        "best": "claude-opus-4-5",
+        "cheap": "gpt-4o-mini",
+        "local": "llama3",
+        "code": "claude-sonnet-4",
+        "vision": "gpt-4o",
+        "long-context": "gemini-1.5-pro"
+      },
+
+      "fallbackChains": {
+        "claude-opus-4-5": ["claude-sonnet-4", "gpt-4o", "gemini-1.5-pro"],
+        "claude-sonnet-4": ["gpt-4o", "gemini-2.0-flash", "claude-3-5-haiku"],
+        "claude-3-5-haiku": ["gpt-4o-mini", "gemini-2.0-flash", "llama3"],
+        "gpt-4o": ["claude-sonnet-4", "gemini-1.5-pro", "gpt-4-turbo"],
+        "gpt-4o-mini": ["claude-3-5-haiku", "gemini-2.0-flash", "gpt-3.5-turbo"],
+        "gemini-1.5-pro": ["claude-sonnet-4", "gpt-4o", "gemini-2.0-flash"],
+        "default": ["claude-sonnet-4", "gpt-4o", "gemini-2.0-flash", "llama3"]
+      },
+
+      "defaultFallbackChain": ["claude-sonnet-4", "gpt-4o", "gemini-2.0-flash"],
+
+      "loadBalancing": {
+        "enabled": true,
+        "strategy": "least_loaded",
+        "healthWeight": 0.3,
+        "latencyWeight": 0.3,
+        "costWeight": 0.2,
+        "successRateWeight": 0.2
+      },
+
+      "costOptimization": {
+        "enabled": true,
+        "preferCheaper": true,
+        "maxCostPerRequest": null,
+        "dailyBudget": null,
+        "alertThreshold": 0.8
+      }
+    },
+
+    "healthCheck": {
+      "enabled": true,
+      "intervalMs": 30000,
+      "timeoutMs": 10000,
+      "unhealthyThreshold": 3,
+      "healthyThreshold": 1,
+      "endpoints": {
+        "claude": "/v1/messages",
+        "openai": "/v1/models",
+        "gemini": "/v1/models",
+        "local": "/api/tags",
+        "azure": "/openai/deployments"
+      }
+    },
+
+    "rateLimiting": {
+      "enabled": true,
+      "strategy": "token_bucket",
+      "globalLimits": {
+        "requestsPerMinute": 100,
+        "tokensPerMinute": 500000,
+        "concurrent": 20
+      },
+      "perProviderLimits": true,
+      "perModelLimits": true,
+      "queueOverflow": "reject",
+      "queueMaxSize": 100,
+      "queueTimeoutMs": 30000
+    },
+
+    "caching": {
+      "enabled": true,
+      "strategy": "semantic",
+      "ttl": 3600000,
+      "maxSize": 1000,
+      "minSimilarity": 0.95,
+      "excludeModels": ["claude-opus-4-5"],
+      "excludeTaskTypes": ["creative", "reasoning"]
+    },
+
+    "retry": {
+      "enabled": true,
+      "maxRetries": 3,
+      "initialDelayMs": 1000,
+      "maxDelayMs": 30000,
+      "backoffMultiplier": 2,
+      "retryableErrors": [
+        "rate_limit_exceeded",
+        "server_error",
+        "timeout",
+        "service_unavailable"
+      ],
+      "nonRetryableErrors": [
+        "invalid_api_key",
+        "content_policy_violation",
+        "invalid_request"
+      ]
+    },
+
+    "logging": {
+      "enabled": true,
+      "level": "info",
+      "includePrompts": false,
+      "includeResponses": false,
+      "includeTokenCounts": true,
+      "includeCosts": true,
+      "includeLatency": true,
+      "redactApiKeys": true
+    },
+
+    "metrics": {
+      "enabled": true,
+      "collectUsage": true,
+      "collectLatency": true,
+      "collectCosts": true,
+      "collectErrors": true,
+      "historySize": 10000,
+      "aggregationIntervalMs": 60000,
+      "exportFormat": "prometheus"
+    },
+
+    "events": {
+      "enabled": true,
+      "emit": [
+        "provider.registered",
+        "provider.unregistered",
+        "provider.health_changed",
+        "provider.status_changed",
+        "llm.completion",
+        "llm.stream_start",
+        "llm.stream_end",
+        "llm.error",
+        "llm.fallback",
+        "llm.rate_limited",
+        "router.rule_matched",
+        "cost.threshold_exceeded"
+      ]
+    },
+
+    "taskTypeMapping": {
+      "code_generation": {
+        "preferredModels": ["claude-sonnet-4", "gpt-4o"],
+        "requiredCapabilities": ["tools"],
+        "maxCost": null,
+        "temperature": 0.2
+      },
+      "code_review": {
+        "preferredModels": ["claude-sonnet-4", "gpt-4o"],
+        "requiredCapabilities": ["tools"],
+        "maxCost": null,
+        "temperature": 0.1
+      },
+      "documentation": {
+        "preferredModels": ["claude-sonnet-4", "gpt-4o"],
+        "requiredCapabilities": [],
+        "maxCost": null,
+        "temperature": 0.3
+      },
+      "analysis": {
+        "preferredModels": ["claude-opus-4-5", "gpt-4o"],
+        "requiredCapabilities": [],
+        "maxCost": null,
+        "temperature": 0.2
+      },
+      "conversation": {
+        "preferredModels": ["claude-sonnet-4", "gpt-4o"],
+        "requiredCapabilities": [],
+        "maxCost": null,
+        "temperature": 0.7
+      },
+      "summarization": {
+        "preferredModels": ["claude-3-5-haiku", "gpt-4o-mini"],
+        "requiredCapabilities": [],
+        "maxCost": 0.01,
+        "temperature": 0.3
+      },
+      "translation": {
+        "preferredModels": ["gpt-4o", "claude-sonnet-4"],
+        "requiredCapabilities": [],
+        "maxCost": null,
+        "temperature": 0.2
+      },
+      "creative": {
+        "preferredModels": ["claude-opus-4-5", "gpt-4o"],
+        "requiredCapabilities": [],
+        "maxCost": null,
+        "temperature": 0.9
+      },
+      "reasoning": {
+        "preferredModels": ["claude-opus-4-5", "gpt-4o"],
+        "requiredCapabilities": [],
+        "maxCost": null,
+        "temperature": 0.1
+      },
+      "quick_task": {
+        "preferredModels": ["claude-3-5-haiku", "gpt-4o-mini", "gemini-2.0-flash"],
+        "requiredCapabilities": [],
+        "maxCost": 0.001,
+        "temperature": 0.5
+      }
+    },
+
+    "unifiedAPI": {
+      "responseNormalization": true,
+      "tokenCountingMethod": "provider_native",
+      "streamingBufferSize": 1024,
+      "defaultTemperature": 0.7,
+      "defaultMaxTokens": 4096,
+      "includeMetadata": true,
+      "timeoutMs": 120000
+    }
+  }
+}

package/templates/multi-project-memory-config.json

@@ -0,0 +1,145 @@
+{
+  "multiProjectMemory": {
+    "enabled": true,
+    "version": "1.0.0",
+
+    "hub": {
+      "storagePath": "~/.elsabro/shared-memory",
+      "maxEntries": 10000,
+      "maxNamespaces": 100,
+      "defaultNamespace": "default",
+      "enableSemanticSearch": true,
+      "embeddingModel": "text-embedding-ada-002",
+      "autoBackup": true,
+      "backupInterval": 86400000
+    },
+
+    "namespaces": {
+      "default": {
+        "visibility": "private",
+        "description": "Personal knowledge base"
+      },
+      "team": {
+        "visibility": "team",
+        "description": "Shared team knowledge"
+      }
+    },
+
+    "entryTypes": {
+      "pattern": {
+        "description": "Reusable code patterns and architectures",
+        "icon": "📐",
+        "autoExtract": true
+      },
+      "mistake": {
+        "description": "Common mistakes to avoid",
+        "icon": "⚠️",
+        "autoExtract": true
+      },
+      "preference": {
+        "description": "Project and coding preferences",
+        "icon": "⭐",
+        "autoExtract": true
+      },
+      "snippet": {
+        "description": "Reusable code snippets",
+        "icon": "📝",
+        "autoExtract": false
+      },
+      "architecture": {
+        "description": "Architecture decisions and diagrams",
+        "icon": "🏗️",
+        "autoExtract": false
+      },
+      "dependency": {
+        "description": "Dependency configurations and notes",
+        "icon": "📦",
+        "autoExtract": true
+      }
+    },
+
+    "linker": {
+      "autoDiscoverProjects": true,
+      "discoveryPaths": [
+        "~/projects",
+        "~/work",
+        "~/code"
+      ],
+      "maxLinks": 50,
+      "defaultPermissions": {
+        "canRead": true,
+        "canWrite": false,
+        "canDelete": false,
+        "canShare": true
+      }
+    },
+
+    "sync": {
+      "enabled": true,
+      "autoSync": false,
+      "syncInterval": 3600000,
+      "conflictResolution": "prompt",
+      "maxRetries": 3,
+      "retryDelay": 5000,
+      "webhooks": {
+        "enabled": false,
+        "timeout": 10000,
+        "retryOnFailure": true
+      }
+    },
+
+    "search": {
+      "minRelevance": 0.3,
+      "maxResults": 50,
+      "useSemanticSearch": true,
+      "semanticWeight": 0.6,
+      "textWeight": 0.4,
+      "boostRecentUsage": true,
+      "usageBoostFactor": 0.1
+    },
+
+    "extraction": {
+      "enabled": true,
+      "autoExtract": true,
+      "minConfidence": 0.7,
+      "sources": {
+        "code": true,
+        "conversation": true,
+        "commit": true,
+        "review": true
+      },
+      "patterns": {
+        "correction": "no,?\\s+(use|do|prefer|always|never)",
+        "teaching": "in this project,?\\s+we\\s+(always|never|prefer)",
+        "reminder": "remember to\\s+|don't forget\\s+",
+        "pattern": "the pattern here is\\s+"
+      }
+    },
+
+    "cleanup": {
+      "enabled": true,
+      "removeUnused": true,
+      "unusedThreshold": 180,
+      "archiveOld": true,
+      "archiveAfterDays": 365,
+      "maxVersions": 10
+    },
+
+    "privacy": {
+      "excludePatterns": [
+        "**/.env*",
+        "**/secrets/**",
+        "**/credentials/**"
+      ],
+      "anonymizeByDefault": false,
+      "requireExplicitShare": true
+    },
+
+    "ui": {
+      "showDashboard": true,
+      "showSuggestions": true,
+      "maxSuggestions": 5,
+      "highlightMatches": true
+    }
+  }
+}