npm - @jsonstudio/llms - Versions diffs - 0.6.34 → 0.6.54 - Mend

@jsonstudio/llms 0.6.34 → 0.6.54

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.

Files changed (38) hide show

package/dist/conversion/codecs/gemini-openai-codec.js CHANGED Viewed

@@ -262,7 +262,6 @@ export function buildOpenAIChatFromGeminiResponse(payload) {
     catch {
         // best-effort policy execution
     }
-    const normalizedFinishReason = toolCalls.length ? 'tool_calls' : finish_reason;
     const chatResp = {
         id: payload?.id || `chatcmpl_${Date.now()}`,
         object: 'chat.completion',
@@ -270,7 +269,7 @@ export function buildOpenAIChatFromGeminiResponse(payload) {
         choices: [
             {
                 index: 0,
-                finish_reason: normalizedFinishReason,
+                finish_reason,
                 message: chatMsg
             }
         ]

package/dist/conversion/compat/profiles/chat-glm.json ADDED Viewed

@@ -0,0 +1,17 @@
+{
+    "id": "chat:glm",
+    "protocol": "openai-chat",
+    "direction": "request",
+    "mappings": [
+        {
+            "action": "rename",
+            "from": "response_format",
+            "to": "metadata.generation.response_format"
+        },
+        {
+            "action": "remove",
+            "path": "metadata.clientModelId"
+        }
+    ],
+    "filters": []
+}

package/dist/conversion/compat/profiles/chat-iflow.json ADDED Viewed

@@ -0,0 +1,36 @@
+{
+    "id": "chat:iflow",
+    "protocol": "openai-chat",
+    "request": {
+        "mappings": [
+            {
+                "action": "remove",
+                "path": "metadata.toolCallIdStyle"
+            },
+            {
+                "action": "remove",
+                "path": "metadata.clientModelId"
+            },
+            {
+                "action": "remove",
+                "path": "metadata.providerHint"
+            }
+        ]
+    },
+    "response": {
+        "mappings": [
+            {
+                "action": "rename",
+                "from": "created_at",
+                "to": "created"
+            },
+            {
+                "action": "convert_responses_output_to_choices"
+            },
+            {
+                "action": "stringify",
+                "path": "choices[*].message.tool_calls[*].function.arguments"
+            }
+        ]
+    }
+}

package/dist/conversion/compat/profiles/chat-lmstudio.json ADDED Viewed

@@ -0,0 +1,37 @@
+{
+    "id": "chat:lmstudio",
+    "protocol": "openai-chat",
+    "request": {
+        "mappings": [
+            {
+                "action": "normalize_tool_choice",
+                "path": "tool_choice",
+                "objectReplacement": "required"
+            }
+        ]
+    },
+    "response": {
+        "mappings": [
+            {
+                "action": "set_default",
+                "path": "object",
+                "value": "chat.completion"
+            },
+            {
+                "action": "set_default",
+                "path": "id",
+                "valueSource": "chat_completion_id"
+            },
+            {
+                "action": "set_default",
+                "path": "created",
+                "valueSource": "timestamp_seconds"
+            },
+            {
+                "action": "set_default",
+                "path": "model",
+                "value": "unknown"
+            }
+        ]
+    }
+}

package/dist/conversion/compat/profiles/chat-qwen.json ADDED Viewed

@@ -0,0 +1,18 @@
+{
+    "id": "chat:qwen",
+    "protocol": "openai-chat",
+    "request": {
+        "mappings": [
+            {
+                "action": "parse_json",
+                "path": "messages[*].tool_calls[*].function.arguments",
+                "fallback": {}
+            },
+            {
+                "action": "stringify",
+                "path": "messages[*].tool_calls[*].function.arguments",
+                "fallback": {}
+            }
+        ]
+    }
+}

package/dist/conversion/compat/profiles/responses-c4m.json ADDED Viewed

@@ -0,0 +1,45 @@
+{
+    "id": "responses:c4m",
+    "protocol": "openai-responses",
+    "request": {
+        "mappings": [
+            {
+                "action": "remove",
+                "path": "max_tokens"
+            },
+            {
+                "action": "remove",
+                "path": "maxTokens"
+            },
+            {
+                "action": "remove",
+                "path": "max_output_tokens"
+            },
+            {
+                "action": "remove",
+                "path": "maxOutputTokens"
+            },
+            {
+                "action": "inject_instruction",
+                "sourcePath": "instructions",
+                "targetPath": "input",
+                "role": "system",
+                "contentType": "input_text",
+                "stripHtml": true,
+                "maxLengthEnv": [
+                    "ROUTECODEX_C4M_INSTRUCTIONS_MAX",
+                    "RCC_C4M_INSTRUCTIONS_MAX",
+                    "ROUTECODEX_COMPAT_INSTRUCTIONS_MAX"
+                ]
+            }
+        ]
+    },
+    "response": {
+        "filters": [
+            {
+                "action": "rate_limit_text",
+                "needle": "The Codex-For.ME service is available, but you have reached the request limit"
+            }
+        ]
+    }
+}

package/dist/conversion/config/compat-profiles.json ADDED Viewed

@@ -0,0 +1,38 @@
+{
+    "version": "1.0.0",
+    "profiles": {
+        "chat:iflow": {
+            "protocol": "openai-chat",
+            "direction": "request",
+            "mappings": [
+                {
+                    "action": "remove",
+                    "path": "messages[*].content[*].annotations"
+                }
+            ],
+            "filters": []
+        },
+        "chat:glm": {
+            "protocol": "openai-chat",
+            "direction": "request",
+            "mappings": [
+                {
+                    "action": "rename",
+                    "from": "response_format",
+                    "to": "extra.response_format"
+                }
+            ],
+            "filters": []
+        },
+        "responses:c4m": {
+            "protocol": "openai-responses",
+            "direction": "response",
+            "filters": [
+                {
+                    "action": "rate_limit_text",
+                    "needle": "The Codex-For.ME service is available, but you have reached the request limit"
+                }
+            ]
+        }
+    }
+}

package/dist/conversion/config/sample-config.json ADDED Viewed

@@ -0,0 +1,314 @@
+{
+    "version": "1.0.0",
+    "description": "LLM Switch Conversion 配置文件",
+    "lastModified": "2024-11-13T00:00:00.000Z",
+    "httpserver": {
+        "port": 5520,
+        "host": "0.0.0.0",
+        "timeout": 30000,
+        "cors": {
+            "enabled": true,
+            "origin": "*",
+            "methods": ["GET", "POST", "PUT", "DELETE", "OPTIONS"]
+        }
+    },
+    "virtualrouter": {
+        "inputProtocol": "openai",
+        "outputProtocol": "openai",
+        "providers": {
+            "chat-provider-1": {
+                "id": "chat-provider-1",
+                "enabled": true,
+                "type": "openai",
+                "name": "Generic Chat Provider",
+                "description": "Example chat-completions provider",
+                "baseUrl": "https://api.example.com/chat",
+                "apiVersion": "v1",
+                "compatibility": {
+                    "type": "openai",
+                    "features": ["streaming", "function-calling"]
+                },
+                "models": {
+                    "chat-model-a": {
+                        "displayName": "Chat-Model-A",
+                        "description": "Latest chat model with improved reasoning",
+                        "supportsStreaming": true,
+                        "maxTokens": 128000,
+                        "contextWindow": 128000,
+                        "capabilities": ["text", "code", "reasoning", "tool-use"],
+                        "pricing": {
+                            "input": "0.001",
+                            "output": "0.002",
+                            "currency": "CNY"
+                        },
+                        "compatibility": {
+                            "type": "openai",
+                            "moduleConfig": {
+                                "policy": {
+                                    "preflight": true
+                                }
+                            }
+                        }
+                    },
+                    "chat-model-b": {
+                        "displayName": "Chat-Model-B",
+                        "description": "Lightweight chat model for faster responses",
+                        "supportsStreaming": true,
+                        "maxTokens": 8192,
+                        "contextWindow": 8192,
+                        "capabilities": ["text", "code", "basic-tool-use"],
+                        "pricing": {
+                            "input": "0.0005",
+                            "output": "0.001",
+                            "currency": "CNY"
+                        },
+                        "compatibility": {
+                            "type": "openai",
+                            "moduleConfig": {
+                                "policy": {
+                                    "preflight": false
+                                }
+                            }
+                        }
+                    },
+                    "chat-model-c": {
+                        "displayName": "Chat-Model-C",
+                        "description": "Ultra-fast chat model for quick responses",
+                        "supportsStreaming": false,
+                        "maxTokens": 4096,
+                        "contextWindow": 4096,
+                        "capabilities": ["text", "code"],
+                        "pricing": {
+                            "input": "0.0002",
+                            "output": "0.0004",
+                            "currency": "CNY"
+                        },
+                        "compatibility": {
+                            "type": "openai",
+                            "moduleConfig": {
+                                "policy": {
+                                    "preflight": false
+                                }
+                            }
+                        }
+                    }
+                },
+                "apiKey": [
+                    "your-api-key-here"
+                ],
+                "auth": {
+                    "type": "apikey",
+                    "apiKey": "your-api-key-here"
+                },
+                "rateLimit": {
+                    "requestsPerMinute": 60,
+                    "tokensPerMinute": 6000,
+                    "concurrent": 5
+                },
+                "healthCheck": {
+                    "enabled": true,
+                    "endpoint": "/health",
+                    "interval": 30000
+                }
+            },
+            "openai": {
+                "id": "openai",
+                "enabled": false,
+                "type": "openai",
+                "name": "OpenAI Provider",
+                "description": "Official OpenAI provider",
+                "baseUrl": "https://api.openai.com/v1",
+                "apiVersion": "v1",
+                "models": {
+                    "gpt-4": {
+                        "displayName": "GPT-4",
+                        "supportsStreaming": true,
+                        "maxTokens": 8192,
+                        "contextWindow": 8192
+                    },
+                    "gpt-4-turbo": {
+                        "displayName": "GPT-4 Turbo",
+                        "supportsStreaming": true,
+                        "maxTokens": 4096,
+                        "contextWindow": 4096
+                    },
+                    "gpt-3.5-turbo": {
+                        "displayName": "GPT-3.5 Turbo",
+                        "supportsStreaming": true,
+                        "maxTokens": 4096,
+                        "contextWindow": 16384
+                    }
+                },
+                "apiKey": [],
+                "auth": {
+                    "type": "apikey"
+                },
+                "rateLimit": {
+                    "requestsPerMinute": 3500,
+                    "tokensPerMinute": 90000,
+                    "concurrent": 100
+                }
+            },
+            "anthropic": {
+                "id": "anthropic",
+                "enabled": false,
+                "type": "anthropic",
+                "name": "Anthropic Claude Provider",
+                "description": "Anthropic Claude provider",
+                "baseUrl": "https://api.anthropic.com",
+                "apiVersion": "2023-06-01",
+                "models": {
+                    "claude-3-sonnet": {
+                        "displayName": "Claude 3 Sonnet",
+                        "supportsStreaming": true,
+                        "maxTokens": 200000,
+                        "contextWindow": 200000
+                    },
+                    "claude-3-haiku": {
+                        "displayName": "Claude 3 Haiku",
+                        "supportsStreaming": false,
+                        "maxTokens": 200000,
+                        "contextWindow": 200000
+                    }
+                },
+                "apiKey": [],
+                "auth": {
+                    "type": "apikey"
+                },
+                "rateLimit": {
+                    "requestsPerMinute": 1000,
+                    "tokensPerMinute": 40000,
+                    "concurrent": 5
+                }
+            }
+        },
+        "routing": {
+            "default": ["chat-provider-1.chat-model-a"],
+            "coding": ["chat-provider-1.chat-model-a"],
+            "longcontext": ["chat-provider-1.chat-model-a"],
+            "thinking": ["chat-provider-1.chat-model-a", "chat-model-b"],
+            "fast": ["chat-model-c"],
+            "vision": [],
+            "websearch": [],
+            "background": [],
+            "fallback": ["chat-provider-1.chat-model-b"]
+        },
+        "rules": {
+            "inputValidation": {
+                "enabled": true,
+                "strict": false,
+                "maxMessageLength": 100000,
+                "maxToolCount": 50
+            },
+            "outputValidation": {
+                "enabled": true,
+                "strict": false,
+                "maxResponseLength": 100000
+            },
+            "contentFiltering": {
+                "enabled": true,
+                "blockedWords": [],
+                "blockedPatterns": []
+            },
+            "rateLimiting": {
+                "enabled": true,
+                "perUser": {
+                    "requestsPerMinute": 10,
+                    "tokensPerMinute": 1000
+                },
+                "perProvider": {
+                    "chat-provider-1": { "requestsPerMinute": 60, "tokensPerMinute": 6000 },
+                    "openai": { "requestsPerMinute": 3500, "tokensPerMinute": 90000 },
+                    "anthropic": { "requestsPerMinute": 1000, "tokensPerMinute": 40000 }
+                }
+            }
+        }
+    },
+    "serverTools": {
+        "replace": {
+            "web_fetch": {
+                "enabled": true,
+                "timeout": 30000,
+                "maxResponseSize": 1000000,
+                "allowedDomains": ["*"],
+                "blockedDomains": []
+            }
+        },
+        "custom": {
+            "enabled": false,
+            "tools": []
+        },
+        "enabled": true
+    },
+    "monitoring": {
+        "enabled": true,
+        "metrics": {
+            "requestCount": true,
+            "responseTime": true,
+            "errorRate": true,
+            "tokenUsage": true,
+            "providerPerformance": true
+        },
+        "logging": {
+            "level": "info",
+            "maxFileSize": "10MB",
+            "rotate": true,
+            "keepFiles": 7
+        },
+        "alerts": {
+            "errorRateThreshold": 0.05,
+            "responseTimeThreshold": 5000,
+            "tokenUsageThreshold": 10000
+        }
+    },
+    "conversion": {
+        "defaultVersion": "v2",
+        "passthrough": {
+            "enabled": true,
+            "criteria": {
+                "sameProtocol": true,
+                "sameProvider": true,
+                "minimalProcessing": true
+            }
+        },
+        "optimization": {
+            "enabled": true,
+            "caching": {
+                "enabled": true,
+                "maxSize": "100MB",
+                "ttl": 300000
+            },
+            "batching": {
+                "enabled": true,
+                "maxBatchSize": 10,
+                "maxBatchWaitTime": 100
+            }
+        }
+    },
+    "security": {
+        "apiKeyEncryption": {
+            "enabled": true,
+            "algorithm": "AES-256-GCM"
+        },
+        "requestValidation": {
+            "enabled": true,
+            "maxRequestSize": "10MB",
+            "allowedOrigins": ["*"],
+            "rateLimit": {
+                "enabled": true,
+                "requestsPerMinute": 1000,
+                "tokensPerMinute": 50000
+            }
+        },
+        "contentFiltering": {
+            "enabled": false,
+            "blockedContent": {
+                "profanity": false,
+                "personalData": false,
+                "malware": true,
+                "spam": true
+            },
+            "customFilters": []
+        }
+    }
+}

package/dist/conversion/config/version-switch.json ADDED Viewed

@@ -0,0 +1,150 @@
+{
+    "defaultVersion": "v2",
+    "envVarName": "LLMSWITCH_CONVERTER_VERSION",
+    "allowRuntimeSwitch": true,
+    "versionPriority": {
+        "v2": 1,
+        "v3": 2,
+        "auto": 0
+    },
+    "versionSettings": {
+        "v2": {
+            "description": "Legacy conversion-core implementation",
+            "stability": "stable",
+            "features": [
+                "proven_stability",
+                "comprehensive_coverage",
+                "extensive_testing"
+            ],
+            "performance": "baseline",
+            "recommendedFor": [
+                "production",
+                "legacy_compatibility",
+                "maximum_stability"
+            ]
+        },
+        "v3": {
+            "description": "New node-based pipeline architecture",
+            "stability": "beta",
+            "features": [
+                "intelligent_passthrough",
+                "node_based_processing",
+                "automatic_optimization",
+                "zero_config"
+            ],
+            "performance": "optimized",
+            "recommendedFor": [
+                "development",
+                "performance_critical",
+                "new_features"
+            ]
+        },
+        "auto": {
+            "description": "Automatically select best version based on request",
+            "stability": "dynamic",
+            "features": [
+                "smart_selection",
+                "performance_optimization",
+                "adaptive_behavior"
+            ],
+            "performance": "variable",
+            "recommendedFor": [
+                "testing",
+                "benchmarking",
+                "a_b_testing"
+            ]
+        }
+    },
+    "switchingRules": {
+        "autoSelection": {
+            "simpleRequests": {
+                "conditions": {
+                    "messageCount": "<= 5",
+                    "toolCount": 0,
+                    "streaming": false
+                },
+                "preferredVersion": "v3",
+                "reason": "Simple requests benefit from v3 passthrough optimization"
+            },
+            "complexRequests": {
+                "conditions": {
+                    "messageCount": "> 20",
+                    "toolCount": "> 10",
+                    "hasStreaming": true
+                },
+                "preferredVersion": "v2",
+                "reason": "Complex requests benefit from v2 stability"
+            },
+            "toolHeavyRequests": {
+                "conditions": {
+                    "toolCount": "> 5",
+                    "hasComplexTools": true
+                },
+                "preferredVersion": "v3",
+                "reason": "V3 has better tool processing architecture"
+            }
+        },
+        "fallbackRules": {
+            "v3FallbackToV2": {
+                "conditions": [
+                    "conversion_error",
+                    "timeout",
+                    "unsupported_endpoint"
+                ],
+                "retryCount": 3
+            },
+            "v2FallbackToV3": {
+                "conditions": [
+                    "performance_degradation",
+                    "memory_pressure"
+                ],
+                "retryCount": 1
+            }
+        }
+    },
+    "monitoring": {
+        "enabled": true,
+        "metrics": [
+            "conversion_success_rate",
+            "average_latency",
+            "passthrough_rate",
+            "error_rate_by_version",
+            "performance_comparison"
+        ],
+        "alertThresholds": {
+            "errorRate": 0.05,
+            "latencyIncrease": 2.0,
+            "successRateDrop": 0.1
+        }
+    },
+    "hooks": {
+        "beforeSwitch": {
+            "logLevel": "info",
+            "metrics": true
+        },
+        "afterSwitch": {
+            "logLevel": "info",
+            "metrics": true,
+            "validate": true
+        },
+        "onError": {
+            "logLevel": "error",
+            "metrics": true,
+            "escalate": true
+        }
+    },
+    "testing": {
+        "enabled": false,
+        "canaryDeployment": {
+            "percentage": 0.1,
+            "version": "v3"
+        },
+        "abTesting": {
+            "enabled": false,
+            "split": {
+                "v2": 0.5,
+                "v3": 0.5
+            }
+        }
+    }
+}

package/dist/conversion/hub/pipeline/compat/compat-engine.d.ts ADDED Viewed

@@ -0,0 +1,4 @@
+import type { JsonObject } from '../../types/json.js';
+import type { CompatApplicationResult } from './compat-types.js';
+export declare function applyRequestCompat(profileId: string | undefined, payload: JsonObject): CompatApplicationResult;
+export declare function applyResponseCompat(profileId: string | undefined, payload: JsonObject): CompatApplicationResult;