hack/tools/scripts/llm-key-hunter/results/_cache/content/chenxingqiang_claude-code-open.json

{"repo": "chenxingqiang/claude-code-open", "files": {"claude-llm-gateway/config/providers.json": {"12f35419acce749c55d89e1881029c7d541687d5": {"t": 1785674676.3370037, "n": 26113, "h": "6760e6729b15828b", "c": "{\n  \"generated_at\": \"2026-01-03T10:11:50.351Z\",\n  \"llm_interface_version\": \"2.0.1495\",\n  \"total_providers\": 5,\n  \"enabled_providers\": 0,\n  \"providers\": {\n    \"generated_at\": \"2026-01-03T08:56:14.035Z\",\n    \"llm_interface_version\": \"2.0.1495\",\n    \"total_providers\": 45,\n    \"enabled_providers\": 7,\n    \"providers\": {\n      \"openai\": {\n        \"enabled\": true,\n        \"priority\": 2,\n        \"models\": [\n          \"gpt-5.2\",\n          \"gpt-5.1\",\n          \"gpt-5\",\n          \"gpt-4.1\",\n          \"gpt-4.1-mini\",\n          \"gpt-4o\",\n          \"gpt-4o-mini\",\n          \"gpt-4-turbo\",\n          \"gpt-4\",\n          \"o4-mini\",\n          \"o3-pro\",\n          \"o3\",\n          \"o3-mini\",\n          \"o1\",\n          \"o1-mini\",\n          \"o1-preview\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true,\n          \"embeddings\": true\n        },\n        \"rate_limit\": 60,\n        \"cost_per_1k_tokens\": 0.03,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.019Z\"\n      },\n      \"anthropic\": {\n        \"enabled\": true,\n        \"priority\": 3,\n        \"models\": [\n          \"claude-opus-4-5-20251215\",\n          \"claude-sonnet-4-5-20251101\",\n          \"claude-haiku-4-5-20251101\",\n          \"claude-opus-4-20250522\",\n          \"claude-sonnet-4-20250514\",\n          \"claude-3-7-sonnet-20250219\",\n          \"claude-3-5-sonnet-20241022\",\n          \"claude-3-5-haiku-20241022\",\n          \"claude-3-opus-20240229\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 50,\n        \"cost_per_1k_tokens\": 0.015,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.020Z\"\n      },\n      \"google\": {\n        \"enabled\": true,\n        \"priority\": 4,\n        \"models\": [\n          \"gemini-3-pro-preview\",\n          \"gemini-2.5-pro\",\n          \"gemini-2.5-pro-preview\",\n          \"gemini-2.5-flash\",\n          \"gemini-2.5-flash-preview\",\n          \"gemini-2.0-flash\",\n          \"gemini-2.0-flash-exp\",\n          \"gemini-1.5-pro\",\n          \"gemini-1.5-flash\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true,\n          \"vision\": true\n        },\n        \"rate_limit\": 100,\n        \"cost_per_1k_tokens\": 0.001,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.020Z\"\n      },\n      \"cohere\": {\n        \"enabled\": false,\n        \"priority\": 6,\n        \"models\": [\n          \"command-r-plus\",\n          \"command-r\",\n          \"command-r-08-2024\",\n          \"command-light\",\n          \"command-nightly\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true,\n          \"embeddings\": true\n        },\n        \"rate_limit\": 40,\n        \"cost_per_1k_tokens\": 0.02,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.021Z\"\n      },\n      \"huggingface\": {\n        \"enabled\": false,\n        \"priority\": 9,\n        \"models\": [\n          \"meta-llama/Llama-4-Scout-17B-16E-Instruct\",\n          \"meta-llama/Llama-4-Maverick-17B-128E-Instruct\",\n          \"meta-llama/Llama-3.3-70B-Instruct\",\n          \"Qwen/Qwen2.5-72B-Instruct\",\n          \"Qwen/QwQ-32B-Preview\",\n          \"deepseek-ai/DeepSeek-V3\",\n          \"mistralai/Mistral-Large-Instruct-2411\",\n          \"microsoft/Phi-4\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.001,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": false,\n        \"last_updated\": \"2026-01-03T08:56:14.021Z\"\n      },\n      \"ollama\": {\n        \"enabled\": true,\n        \"priority\": 5,\n        \"models\": [\n          \"llama4:scout\",\n          \"llama4:maverick\",\n          \"llama3.3:70b\",\n          \"llama3.2:3b\",\n          \"llama3.1:8b\",\n          \"qwen2.5:72b\",\n          \"qwen2.5:32b\",\n          \"qwen2.5:7b\",\n          \"deepseek-v3:latest\",\n          \"deepseek-r1:70b\",\n          \"deepseek-r1:32b\",\n          \"deepseek-r1:8b\",\n          \"gemma2:27b\",\n          \"gemma2:9b\",\n          \"mistral:7b\",\n          \"codellama:34b\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 1000,\n        \"cost_per_1k_tokens\": 0.001,\n        \"requires_api_key\": false,\n        \"local\": true,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"mistral\": {\n        \"enabled\": true,\n        \"priority\": 7,\n        \"models\": [\n          \"mistral-large-3\",\n          \"mistral-large-2411\",\n          \"mistral-large-latest\",\n          \"mistral-medium-latest\",\n          \"mistral-small-latest\",\n          \"mistral-saba-24b\",\n          \"codestral-2501\",\n          \"codestral-latest\",\n          \"pixtral-large-latest\",\n          \"pixtral-12b-2409\",\n          \"open-mixtral-8x22b\",\n          \"open-mixtral-8x7b\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 50,\n        \"cost_per_1k_tokens\": 0.025,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"groq\": {\n        \"enabled\": true,\n        \"priority\": 8,\n        \"models\": [\n          \"llama-4-scout-17b-16e-instruct\",\n          \"llama-4-maverick-17b-128e-instruct\",\n          \"llama-guard-4-12b\",\n          \"llama-3.3-70b-versatile\",\n          \"llama-3.1-70b-versatile\",\n          \"llama-3.1-8b-instant\",\n          \"mistral-saba-24b\",\n          \"mixtral-8x7b-32768\",\n          \"gemma2-9b-it\",\n          \"qwen-qwq-32b\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.001,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"perplexity\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"llama-3.1-sonar-small-128k-online\",\n          \"llama-3.1-sonar-large-128k-online\",\n          \"llama-3.1-sonar-huge-128k-online\",\n          \"llama-3.1-sonar-small-128k-chat\",\n          \"llama-3.1-sonar-large-128k-chat\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 20,\n        \"cost_per_1k_tokens\": 0.02,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"ai21\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"jamba-1.5-large\",\n          \"jamba-1.5-mini\",\n          \"j2-ultra\",\n          \"j2-mid\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 20,\n        \"cost_per_1k_tokens\": 0.025,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"nvidia\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"meta/llama-3.1-405b-instruct\",\n          \"meta/llama-3.1-70b-instruct\",\n          \"nvidia/llama-3.1-nemotron-70b-instruct\",\n          \"mistralai/mixtral-8x22b-instruct-v0.1\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.015,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"fireworks\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"accounts/fireworks/models/llama4-scout-instruct\",\n          \"accounts/fireworks/models/llama4-maverick-instruct\",\n          \"accounts/fireworks/models/llama-v3p3-70b-instruct\",\n          \"accounts/fireworks/models/llama-v3p1-405b-instruct\",\n          \"accounts/fireworks/models/deepseek-v3\",\n          \"accounts/fireworks/models/deepseek-r1\",\n          \"accounts/fireworks/models/qwen2p5-72b-instruct\",\n          \"accounts/fireworks/models/mixtral-8x22b-instruct\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 40,\n        \"cost_per_1k_tokens\": 0.002,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"together\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8\",\n          \"meta-llama/Llama-4-Scout-17B-16E-Instruct\",\n          \"meta-llama/Llama-3.3-70B-Instruct-Turbo\",\n          \"meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo\",\n          \"meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo\",\n          \"deepseek-ai/DeepSeek-V3\",\n          \"deepseek-ai/DeepSeek-R1\",\n          \"Qwen/Qwen2.5-72B-Instruct-Turbo\",\n          \"Qwen/QwQ-32B\",\n          \"mistralai/Mixtral-8x22B-Instruct-v0.1\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 40,\n        \"cost_per_1k_tokens\": 0.002,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"anyscale\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"meta-llama/Llama-3-70b-chat-hf\",\n          \"meta-llama/Llama-3-8b-chat-hf\",\n          \"mistralai/Mixtral-8x7B-Instruct-v0.1\",\n          \"codellama/CodeLlama-70b-Instruct-hf\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 50,\n        \"cost_per_1k_tokens\": 0.001,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"deepseek\": {\n        \"enabled\": true,\n        \"priority\": 1,\n        \"models\": [\n          \"deepseek-chat\",\n          \"deepseek-reasoner\",\n          \"deepseek-v3.2-speciale\",\n          \"deepseek-v3.1\",\n          \"deepseek-v3-0324\",\n          \"deepseek-r1-0528\",\n          \"deepseek-r1\",\n          \"deepseek-coder\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.001,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"replicate\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"meta/llama-3.3-70b-instruct\",\n          \"meta/llama-3.1-405b-instruct\",\n          \"meta/llama-3-70b-instruct\",\n          \"mistralai/mixtral-8x7b-instruct-v0.1\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 20,\n        \"cost_per_1k_tokens\": 0.005,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"gooseai\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"forefront\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"writer\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"neets\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"lamini\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"hyperbee\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"novita\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"shuttle\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"theb\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"corcel\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.034Z\"\n      },\n      \"aimlapi\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.035Z\"\n      },\n      \"ailayer\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.035Z\"\n      },\n      \"monster\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.035Z\"\n      },\n      \"deepinfra\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"meta-llama/Llama-3.3-70B-Instruct-Turbo\",\n          \"meta-llama/Meta-Llama-3.1-405B-Instruct\",\n          \"Qwen/Qwen2.5-72B-Instruct\",\n          \"mistralai/Mixtral-8x22B-Instruct-v0.1\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.035Z\"\n      },\n      \"friendliai\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.035Z\"\n      },\n      \"reka\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.035Z\"\n      },\n      \"voyage\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.035Z\"\n      },\n      \"watsonx\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"default-model\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.035Z\"\n      },\n      \"zhipu\": {\n        \"enabled\": false,\n        \"priority\": 10,\n        \"models\": [\n          \"glm-4-plus\",\n          \"glm-4-0520\",\n          \"glm-4-air\",\n          \"glm-4-airx\",\n          \"glm-4-flash\",\n          \"glm-4v-plus\",\n          \"glm-4v\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true,\n          \"vision\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.035Z\"\n      },\n      \"xai\": {\n        \"enabled\": false,\n        \"priority\": 8,\n        \"models\": [\n          \"grok-4-heavy\",\n          \"grok-4\",\n          \"grok-4-fast\",\n          \"grok-3\",\n          \"grok-3-mini\",\n          \"grok-2-vision\",\n          \"grok-2\",\n          \"grok-beta\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true,\n          \"vision\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.02,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T09:30:00.000Z\"\n      },\n      \"bedrock\": {\n        \"enabled\": false,\n        \"priority\": 9,\n        \"models\": [\n          \"anthropic.claude-opus-4-5-v1:0\",\n          \"anthropic.claude-sonnet-4-5-v1:0\",\n          \"anthropic.claude-haiku-4-5-v1:0\",\n          \"anthropic.claude-opus-4-v1:0\",\n          \"anthropic.claude-sonnet-4-v1:0\",\n          \"anthropic.claude-3-5-sonnet-v2:0\",\n          \"amazon.nova-premier-v1:0\",\n          \"amazon.nova-pro-v1:0\",\n          \"amazon.nova-lite-v1:0\",\n          \"amazon.titan-text-premier-v1:0\",\n          \"meta.llama4-scout-17b-instruct-v1:0\",\n          \"meta.llama4-maverick-17b-instruct-v1:0\",\n          \"meta.llama3-3-70b-instruct-v1:0\",\n          \"mistral.mistral-large-2411-v1:0\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 50,\n        \"cost_per_1k_tokens\": 0.015,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T09:30:00.000Z\"\n      },\n      \"azure\": {\n        \"enabled\": false,\n        \"priority\": 9,\n        \"models\": [\n          \"gpt-4o\",\n          \"gpt-4o-mini\",\n          \"gpt-4-turbo\",\n          \"gpt-4\",\n          \"gpt-35-turbo\",\n          \"o1\",\n          \"o1-mini\",\n          \"o1-preview\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true,\n          \"embeddings\": true\n        },\n        \"rate_limit\": 60,\n        \"cost_per_1k_tokens\": 0.03,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T09:30:00.000Z\"\n      },\n      \"openrouter\": {\n        \"enabled\": false,\n        \"priority\": 7,\n        \"models\": [\n          \"openai/gpt-5.2\",\n          \"openai/gpt-4o\",\n          \"openai/o3\",\n          \"openai/o1\",\n          \"anthropic/claude-opus-4\",\n          \"anthropic/claude-sonnet-4\",\n          \"anthropic/claude-3.5-sonnet\",\n          \"google/gemini-2.5-pro\",\n          \"google/gemini-2.5-flash\",\n          \"meta-llama/llama-4-maverick\",\n          \"meta-llama/llama-4-scout\",\n          \"deepseek/deepseek-v3.1\",\n          \"deepseek/deepseek-r1\",\n          \"mistralai/mistral-large-3\",\n          \"qwen/qwen-2.5-72b-instruct\",\n          \"x-ai/grok-4\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 100,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T09:30:00.000Z\"\n      },\n      \"alibaba\": {\n        \"enabled\": false,\n        \"priority\": 8,\n        \"models\": [\n          \"qwen3-235b-a22b\",\n          \"qwen3-32b\",\n          \"qwen-max\",\n          \"qwen-max-longcontext\",\n          \"qwen-plus\",\n          \"qwen-turbo\",\n          \"qwen2.5-72b-instruct\",\n          \"qwen2.5-coder-32b-instruct\",\n          \"qwq-32b-preview\",\n          \"qwen-vl-max\",\n          \"qwen-vl-plus\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true,\n          \"vision\": true\n        },\n        \"rate_limit\": 60,\n        \"cost_per_1k_tokens\": 0.008,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T09:30:00.000Z\"\n      },\n      \"moonshot\": {\n        \"enabled\": false,\n        \"priority\": 9,\n        \"models\": [\n          \"moonshot-v1-8k\",\n          \"moonshot-v1-32k\",\n          \"moonshot-v1-128k\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.012,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T09:30:00.000Z\"\n      },\n      \"baidu\": {\n        \"enabled\": false,\n        \"priority\": 9,\n        \"models\": [\n          \"ernie-4.0-8k\",\n          \"ernie-4.0-turbo-8k\",\n          \"ernie-3.5-8k\",\n          \"ernie-speed-8k\",\n          \"ernie-lite-8k\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.015,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T09:30:00.000Z\"\n      },\n      \"minimax\": {\n        \"enabled\": false,\n        \"priority\": 9,\n        \"models\": [\n          \"abab6.5s-chat\",\n          \"abab6.5-chat\",\n          \"abab5.5s-chat\",\n          \"abab5.5-chat\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 30,\n        \"cost_per_1k_tokens\": 0.01,\n        \"requires_api_key\": true,\n        \"local\": false,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T09:30:00.000Z\"\n      },\n      \"llamacpp\": {\n        \"enabled\": {},\n        \"priority\": 10,\n        \"models\": [\n          \"llama-3.3-70b-instruct\",\n          \"llama-3.1-8b-instruct\",\n          \"qwen2.5-7b-instruct\",\n          \"mistral-7b-instruct-v0.3\",\n          \"codellama-7b-instruct\"\n        ],\n        \"capabilities\": {\n          \"chat\": true,\n          \"completion\": true\n        },\n        \"rate_limit\": 1000,\n        \"cost_per_1k_tokens\": 0.001,\n        \"requires_api_key\": false,\n        \"local\": true,\n        \"streaming_support\": true,\n        \"last_updated\": \"2026-01-03T08:56:14.035Z\"\n      }\n    }\n  }\n}"}}, "README.md": {"12f35419acce749c55d89e1881029c7d541687d5": {"t": 1785746514.2234704, "n": 15927, "h": "02061420ad35d706", "c": "# 🧠 Claude LLM Gateway - Intelligent Multi-LLM API Gateway\n\n**Intelligent Multi-Model API Gateway - Unified Interface for 36+ Large Language Models**\n\n[![NPM Version](https://img.shields.io/npm/v/claude-llm-gateway)](https://www.npmjs.com/package/claude-llm-gateway)\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)\n[![Node.js](https://img.shields.io/badge/Node.js-16%2B-green)](https://nodejs.org/)\n\nClaude LLM Gateway is an intelligent API gateway that provides a unified multi-model LLM interface for Claude Code and other applications. It features intelligent model selection, cost optimization, load balancing, and failover capabilities.\n\n![Claude LLM Gateway Dashboard](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/dashboard.png)\n\n## ✨ Key Features\n\n- 🧠 **Intelligent Model Selection** - Auto-select optimal models based on task types\n- 🌍 **36+ Model Support** - OpenAI, Anthropic, Google, Meta, DeepSeek, and more\n- 🔄 **Bilingual Interface** - Complete Chinese/English internationalization\n- 💰 **Cost Optimization** - Smart routing to reduce API call costs\n- 📊 **Real-time Monitoring** - Web interface for all provider status monitoring\n- 🛡️ **Failover Protection** - Automatic switching to available backup models\n- 🔐 **Secure Gateway** - API key management and access control\n\n---\n\n## 📋 Table of Contents\n\n1. [🚀 Quick Start](#-quick-start)\n2. [📦 Installation](#-installation)\n3. [⚙️ Configuration](#-configuration)\n4. [🌐 Web Interface](#-web-interface)\n5. [🛠️ CLI Commands](#-cli-commands)\n6. [🔧 API Usage](#-api-usage)\n7. [🌍 Supported Models](#-supported-models)\n8. [📊 Intelligent Selection](#-intelligent-selection)\n9. [🔐 Security Configuration](#-security-configuration)\n10. [📝 Troubleshooting](#-troubleshooting)\n\n---\n\n## 🚀 Quick Start\n\n### System Requirements\n\n- **Node.js** v16+ - [Download here](https://nodejs.org/)\n- **npm** v7+ (comes with Node.js)\n- **Supported OS**: macOS, Linux, Windows\n\n### One-Click Installation\n\n```bash\n# Install globally via NPM\nnpm install -g claude-llm-gateway\n\n# Start immediately\nnpx claude-llm-gateway start --port 8765\n```\n\n🎉 **Done!** Open your browser and visit http://localhost:8765\n\n---\n\n## 📦 Installation\n\n### Method 1: Global Installation (Recommended)\n```bash\n# Install globally\nnpm install -g claude-llm-gateway\n\n# Start service\nclaude-llm-gateway start --port 8765\n\n# Run in background\nclaude-llm-gateway start --port 8765 --daemon\n```\n\n### Method 2: Local Project Installation\n```bash\n# Add to project dependencies\nnpm install claude-llm-gateway\n\n# Run with npx\nnpx claude-llm-gateway start --port 8765\n```\n\n### Method 3: Build from Source\n```bash\n# Clone repository\ngit clone https://github.com/chenxingqiang/claude-code-jailbreak.git\ncd claude-code-jailbreak/claude-llm-gateway\n\n# Install dependencies\nnpm install\n\n# Start development mode\nnpm run dev\n```\n\n## ⚙️ Configuration\n\n### Environment Variable Setup\n\nCreate a `.env` file or set environment variables:\n\n```bash\n# OpenAI\nOPENAI_API_KEY=sk-your-openai-key\n\n# Anthropic Claude\nANTHROPIC_API_KEY=sk-ant-your-anthropic-key\n\n# Google Gemini\nGOOGLE_API_KEY=your-google-api-key\n\n# DeepSeek (Recommended, large free quota)\nDEEPSEEK_API_KEY=sk-your-deepseek-key\n\n# Other optional configurations\nGROQ_API_KEY=gsk_your-groq-key\nMISTRAL_API_KEY=your-mistral-key\nMOONSHOT_API_KEY=sk-your-moonshot-key\n```\n\n### Quick Configuration Template\n\n```bash\n# Copy environment variable template\ncp env.example .env\n\n# Edit configuration file\nnano .env\n```\n\n**🔒 Security Tips**: \n- Never commit API keys to Git repositories\n- Use environment variables for sensitive information\n- Rotate API keys regularly\n\n---\n\n## 🌐 Web Interface\n\nAfter starting the service, you can access the management interface in two ways:\n\n### 🚀 Method 1: Integrated Web Interface (Default)\n```bash\nclaude-llm-gateway start --port 8765\n# Visit: http://localhost:8765\n```\n\n### 🌟 Method 2: Dedicated Web Server (Recommended)\n```bash\n# Start Gateway API service\nclaude-llm-gateway start --port 8765 --daemon\n\n# Start dedicated Web UI server  \ncd claude-llm-gateway && npm run web\n# Visit: http://localhost:9000\n\n# Or start both services together\nnpm run both\n```\n\n**🎯 Benefits of Method 2:**\n- **Better Performance**: Web UI and API separated\n- **Independent Scaling**: Scale web and API separately  \n- **Enhanced Security**: API and UI can run on different networks\n- **Development Friendly**: Hot reload for UI development\n\n### Interface Overview\n\nThe Claude LLM Gateway features a comprehensive web management interface with multiple functional tabs:\n\n#### 📊 Main Dashboard\n![Claude LLM Gateway Dashboard](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/dashboard.png)\n\nThe main dashboard provides real-time status overview with provider health monitoring and system statistics.\n\n#### 🔝 Top Status Bar\n- **Language Toggle**: One-click Chinese/English switching\n- **Refresh Status**: Manually refresh all provider statuses\n- **Last Updated**: Display data update timestamp\n\n#### 📊 Status Overview\n- **Healthy Providers**: Number of currently available models (shown as 2 in the image)\n- **Total Providers**: Total supported provider count (36)\n- **Available Models**: Total number of available models (78)\n- **Total Requests**: Request statistics\n\n#### 📝 Main Tabs\n\n**1. Provider Management** (as shown in image)\n- View real-time status of all LLM providers\n- Green toggle indicates enabled, red dot indicates unhealthy\n- Display request count, response time, cost information for each provider\n- Supported model lists (e.g., OpenAI's gpt-4, gpt-3.5-turbo, etc.)\n\n**2. Configuration** \n![Configuration Interface](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/configs.png)\n\n- Environment variable configuration with secure API key management\n- Real-time gateway settings adjustment\n- Provider-specific configuration options\n\n**3. API Endpoints**\n![API Endpoints](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/api-endpoints.png)\n\n- Complete API documentation with live examples\n- Claude-compatible endpoints with usage instructions\n- Token management API references\n- Health check and provider status interfaces\n\n**4. Intelligent Selection**\n![Model Selection Interface](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/modelsection.png)\n\n- Comprehensive model capability matrix with detailed statistics\n- Performance analytics and optimization insights\n- Task-based intelligent selection strategies\n- Real-time model comparison and recommendations\n\n**5. Real-time Logs**\n![Real-time Logs](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/realtime-logs.png)\n\n- Live system activity monitoring with auto-scroll\n- Detailed request/response logging\n- Provider health status updates\n- System diagnostic information\n\n#### 🎛️ Provider Card Operations\nEach provider card contains:\n- **Status Indicator**: Healthy/unhealthy status\n- **Enable Toggle**: One-click enable/disable provider\n- **Test Button**: Test provider connection\n- **Configure Button**: Quick jump to configuration page\n- **Error Messages**: Display specific error information (e.g., \"API key not configured\")\n\n---\n\n## 🛠️ CLI Commands\n\n### Basic Commands\n\n```bash\n# Start service\nclaude-llm-gateway start [options]\n\n# Run in background (daemon mode)\nclaude-llm-gateway start --daemon\n\n# Stop service\nclaude-llm-gateway stop\n\n# Check status\nclaude-llm-gateway status\n\n# Show help\nclaude-llm-gateway --help\n```\n\n### Web Server Commands\n\n```bash\n# Start dedicated web server (port 9000)\ncd claude-llm-gateway && npm run web\n\n# Start web server in development mode\nnpm run web:dev\n\n# Start both API gateway and web server\nnpm run both\n```\n\n### Startup Options\n\n```bash\n# Specify port (default: 8765)\nclaude-llm-gateway start --port 3000\n\n# Specify host (default: localhost)\nclaude-llm-gateway start --host 0.0.0.0\n\n# Run in background (daemon mode)\nclaude-llm-gateway start --daemon\n\n# Combined usage\nclaude-llm-gateway start --port 8765 --host 0.0.0.0 --daemon\n```\n\n### Status Management\n\n```bash\n# View detailed status\nclaude-llm-gateway status\n\n# Check all provider statuses\ncurl http://localhost:8765/providers\n\n# Health check\ncurl http://localhost:8765/health\n```\n\n---\n\n## 🔧 API Usage\n\n### Claude-Compatible Interface\n\nClaude LLM Gateway provides fully compatible Claude API interfaces:\n\n```bash\n# Claude Messages API\ncurl -X POST http://localhost:8765/v1/messages \\\n  -H \"Content-Type: application/json\" \\\n  -H \"x-api-key: your-api-key\" \\\n  -d '{\n    \"model\": \"claude-3-sonnet-20240229\",\n    \"max_tokens\": 1024,\n    \"messages\": [\n      {\n        \"role\": \"user\", \n        \"content\": \"Hello, world!\"\n      }\n    ]\n  }'\n```\n\n### OpenAI-Compatible Interface\n\nAlso supports OpenAI ChatGPT interface format:\n\n```bash\n# OpenAI Chat Completions API\ncurl -X POST http://localhost:8765/v1/chat/completions \\\n  -H \"Content-Type: application/json\" \\\n  -H \"Authorization: Bearer your-api-key\" \\\n  -d '{\n    \"model\": \"gpt-4\",\n    \"messages\": [\n      {\n        \"role\": \"user\",\n        \"content\": \"Hello, world!\"\n      }\n    ]\n  }'\n```\n\n### Intelligent Routing\n\nWhen no model is specified, the gateway automatically selects the optimal model:\n\n```bash\ncurl -X POST http://localhost:8765/v1/messages \\\n  -H \"Content-Type: application/json\" \\\n  -d '{\n    \"max_tokens\": 1024,\n    \"messages\": [\n      {\n        \"role\": \"user\",\n        \"content\": \"Write a Python function to calculate Fibonacci numbers\"\n      }\n    ]\n  }'\n```\n\n### Node.js SDK Usage\n\n```javascript\nconst { ClaudeLLMGateway } = require('claude-llm-gateway');\n\n// Create gateway instance\nconst gateway = new ClaudeLLMGateway();\n\n// Start service\nawait gateway.start(8765);\n\n// Use built-in client\nconst response = await gateway.chat({\n  messages: [\n    { role: 'user', content: 'Hello!' }\n  ]\n});\n\nconsole.log(response.content);\n```\n\n---\n\n## 🌍 Supported Models\n\n### OpenAI Models\n- **GPT-4o** (gpt-4o) - Latest multimodal model\n- **GPT-4 Turbo** (gpt-4-turbo) - High-performance version\n- **GPT-4** (gpt-4) - Classic version\n- **GPT-4o Mini** (gpt-4o-mini) - Lightweight version\n- **GPT-3.5 Turbo** (gpt-3.5-turbo) - Fast response\n\n### Anthropic Models  \n- **Claude 3 Opus** (claude-3-opus) - Strongest reasoning capabilities\n- **Claude 3.5 Sonnet** (claude-3.5-sonnet) - Balanced performance\n- **Claude 3 Sonnet** (claude-3-sonnet) - Standard version\n- **Claude 3 Haiku** (claude-3-haiku) - Fast response\n\n### Google Models\n- **Gemini Ultra** (gemini-ultra) - Flagship model\n- **Gemini 1.5 Pro** (gemini-1.5-pro) - Long context\n- **Gemini Pro** (gemini-pro) - Standard version\n\n### Meta Models\n- **Llama 3.1 405B** (llama-3.1-405b) - Ultra-large model\n- **Llama 3.1 70B** (llama-3.1-70b) - High performance\n- **Llama 3.1 8B** (llama-3.1-8b) - Lightweight\n\n### Chinese-Optimized Models\n- **DeepSeek V3** (deepseek-v3) - Chinese programming\n- **DeepSeek Chat** (deepseek-chat) - Chinese conversation\n- **DeepSeek Coder** (deepseek-coder) - Code generation\n- **Qianwen Max** (qianwen-max) - Alibaba's large model\n- **GLM-4** (zhipu-glm-4) - Zhipu AI\n- **Moonshot V1** (moonshot-v1) - Dark Side of the Moon\n\n### Other Models\n- **Mistral Large** (mistral-large) - European AI\n- **Groq Models** - Ultra-fast inference\n- **Cohere Command R+** - Enterprise-grade\n- **Grok-1** (grok-1) - xAI humor model\n\n**Total**: 30+ model providers, 78+ available models\n\n---\n\n## 📊 Intelligent Selection\n\n### Task Type Detection\n\nThe gateway can automatically detect task types and select optimal models:\n\n#### 🔧 Programming Tasks\n```\nInput: \"Write a Python function to calculate prime numbers\"\nRecommended: DeepSeek Coder, GPT-4, Claude 3.5 Sonnet\nReason: Strong code generation capabilities, accurate syntax\n```\n\n#### 📝 Creative Tasks  \n```\nInput: \"Write an article about AI\"\nRecommended: Claude 3 Opus, GPT-4, Gemini Pro\nReason: Strong creative abilities, natural writing style\n```\n\n#### 🔍 Analysis Tasks\n```\nInput: \"Analyze the trends in this data report\"\nRecommended: Claude 3 Opus, GPT-4 Turbo, Gemini Ultra  \nReason: Strong logical reasoning, deep analysis\n```\n\n#### 💬 Conversation Tasks\n```\nInput: \"Hello, please introduce yourself\"\nRecommended: GPT-3.5 Turbo, Claude 3 Haiku, DeepSeek Chat\nReason: Fast response, low cost, natural conversation\n```\n\n### Cost Optimization Strategies\n\n- **Intelligent Downgrading**: Automatically use lower-cost models for simple tasks\n- **Parallel Requests**: Send requests to multiple providers simultaneously, choose fastest response\n- **Failover**: Automatically switch to backup models when primary models are unavailable\n- **Load Balancing**: Distribute requests based on provider response time and success rate\n\n## 🔐 Security Configuration\n\n### API Key Management\n\n```bash\n# Recommended: Use environment variables\nexport OPENAI_API_KEY=\"sk-your-key\"\n\n# Avoid: Hard-coding in code\n# const apiKey = \"sk-your-key\"  ❌\n```\n\n### Access Control\n\n```bash\n# Restrict access sources\nclaude-llm-gateway start --host 127.0.0.1  # Local access only\n\n# Production environment\nclaude-llm-gateway start --host 0.0.0.0    # Allow external access\n```\n\n### Security Measures\n\n- ✅ **CORS Configuration**: Prevent cross-origin attacks\n- ✅ **Rate Limiting**: Prevent API abuse  \n- ✅ **Helmet Security Headers**: Enhanced security\n- ✅ **Input Validation**: Prevent injection attacks\n- ✅ **Error Handling**: No sensitive information leakage\n\n---\n\n## 📝 Troubleshooting\n\n### Common Issues\n\n**Q: Port already in use error on startup**\n```bash\n# Check port usage\nlsof -i :8765\n\n# Use different port\nclaude-llm-gateway start --port 8766\n```\n\n**Q: Invalid API key error**\n```bash\n# Check environment variables\necho $OPENAI_API_KEY\n\n# Reset key\nexport OPENAI_API_KEY=\"sk-your-new-key\"\n```\n\n**Q: Cannot connect to a provider**\n```bash\n# Check network connection\ncurl -I https://api.openai.com\n\n# View detailed errors\nclaude-llm-gateway start --verbose\n```\n\n**Q: Intelligent selection not working**\n```bash\n# View model capability configuration\ncurl http://localhost:8765/model-stats\n\n# Check provider status\ncurl http://localhost:8765/providers\n```\n\n### Debug Mode\n\n```bash\n# Enable detailed logging\nexport LOG_LEVEL=debug\nclaude-llm-gateway start\n\n# View real-time logs\ntail -f logs/gateway.log\n```\n\n### Performance Optimization\n\n```bash\n# Increase concurrent connections\nexport MAX_CONNECTIONS=100\n\n# Enable caching\nexport ENABLE_CACHE=true\n\n# Set timeout\nexport REQUEST_TIMEOUT=30000\n```\n\n---\n\n## 🤝 Contributing\n\nWelcome to contribute code! Please follow these steps:\n\n1. **Fork the repository** \n2. **Create feature branch**: `git checkout -b feature/amazing-feature`\n3. **Commit changes**: `git commit -m 'Add amazing feature'`\n4. **Push branch**: `git push origin feature/amazing-feature`\n5. **Create Pull Request**\n\n### Development Environment\n\n```bash\n# Clone repository\ngit clone https://github.com/chenxingqiang/claude-code-jailbreak.git\ncd claude-code-jailbreak/claude-llm-gateway\n\n# Install dependencies\nnpm install\n\n# Run tests\nnpm test\n\n# Start development mode\nnpm run dev\n```\n\n---\n\n## 📄 License\n\nThis project is licensed under the [MIT License](LICENSE)\n\n---\n\n## 🌟 Star History\n\nIf this project helps you, please give us a ⭐!\n\n[![Star History Chart](https://api.star-history.com/svg?repos=chenxingqiang/claude-code-jailbreak&type=Date)](https://star-history.com/#chenxingqiang/claude-code-jailbreak&Date)\n\n---\n\n## 📞 Support & Contact\n\n- 📁 **GitHub**: [claude-code-jailbreak](https://github.com/chenxingqiang/claude-code-jailbreak)\n- 📦 **NPM**: [claude-llm-gateway](https://www.npmjs.com/package/claude-llm-gateway)\n- 🐛 **Issues**: [GitHub Issues](https://github.com/chenxingqiang/claude-code-jailbreak/issues)\n- 💬 **Discussions**: [GitHub Discussions](https://github.com/chenxingqiang/claude-code-jailbreak/discussions)\n\n---\n\n**🎯 Making AI model selection simple and intelligent!**\n"}, "f6c731feac8be62099674cdcc7b4b769700e02ac": {"t": 1785765128.0603755, "n": 15927, "h": "02061420ad35d706", "c": "# 🧠 Claude LLM Gateway - Intelligent Multi-LLM API Gateway\n\n**Intelligent Multi-Model API Gateway - Unified Interface for 36+ Large Language Models**\n\n[![NPM Version](https://img.shields.io/npm/v/claude-llm-gateway)](https://www.npmjs.com/package/claude-llm-gateway)\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)\n[![Node.js](https://img.shields.io/badge/Node.js-16%2B-green)](https://nodejs.org/)\n\nClaude LLM Gateway is an intelligent API gateway that provides a unified multi-model LLM interface for Claude Code and other applications. It features intelligent model selection, cost optimization, load balancing, and failover capabilities.\n\n![Claude LLM Gateway Dashboard](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/dashboard.png)\n\n## ✨ Key Features\n\n- 🧠 **Intelligent Model Selection** - Auto-select optimal models based on task types\n- 🌍 **36+ Model Support** - OpenAI, Anthropic, Google, Meta, DeepSeek, and more\n- 🔄 **Bilingual Interface** - Complete Chinese/English internationalization\n- 💰 **Cost Optimization** - Smart routing to reduce API call costs\n- 📊 **Real-time Monitoring** - Web interface for all provider status monitoring\n- 🛡️ **Failover Protection** - Automatic switching to available backup models\n- 🔐 **Secure Gateway** - API key management and access control\n\n---\n\n## 📋 Table of Contents\n\n1. [🚀 Quick Start](#-quick-start)\n2. [📦 Installation](#-installation)\n3. [⚙️ Configuration](#-configuration)\n4. [🌐 Web Interface](#-web-interface)\n5. [🛠️ CLI Commands](#-cli-commands)\n6. [🔧 API Usage](#-api-usage)\n7. [🌍 Supported Models](#-supported-models)\n8. [📊 Intelligent Selection](#-intelligent-selection)\n9. [🔐 Security Configuration](#-security-configuration)\n10. [📝 Troubleshooting](#-troubleshooting)\n\n---\n\n## 🚀 Quick Start\n\n### System Requirements\n\n- **Node.js** v16+ - [Download here](https://nodejs.org/)\n- **npm** v7+ (comes with Node.js)\n- **Supported OS**: macOS, Linux, Windows\n\n### One-Click Installation\n\n```bash\n# Install globally via NPM\nnpm install -g claude-llm-gateway\n\n# Start immediately\nnpx claude-llm-gateway start --port 8765\n```\n\n🎉 **Done!** Open your browser and visit http://localhost:8765\n\n---\n\n## 📦 Installation\n\n### Method 1: Global Installation (Recommended)\n```bash\n# Install globally\nnpm install -g claude-llm-gateway\n\n# Start service\nclaude-llm-gateway start --port 8765\n\n# Run in background\nclaude-llm-gateway start --port 8765 --daemon\n```\n\n### Method 2: Local Project Installation\n```bash\n# Add to project dependencies\nnpm install claude-llm-gateway\n\n# Run with npx\nnpx claude-llm-gateway start --port 8765\n```\n\n### Method 3: Build from Source\n```bash\n# Clone repository\ngit clone https://github.com/chenxingqiang/claude-code-jailbreak.git\ncd claude-code-jailbreak/claude-llm-gateway\n\n# Install dependencies\nnpm install\n\n# Start development mode\nnpm run dev\n```\n\n## ⚙️ Configuration\n\n### Environment Variable Setup\n\nCreate a `.env` file or set environment variables:\n\n```bash\n# OpenAI\nOPENAI_API_KEY=sk-your-openai-key\n\n# Anthropic Claude\nANTHROPIC_API_KEY=sk-ant-your-anthropic-key\n\n# Google Gemini\nGOOGLE_API_KEY=your-google-api-key\n\n# DeepSeek (Recommended, large free quota)\nDEEPSEEK_API_KEY=sk-your-deepseek-key\n\n# Other optional configurations\nGROQ_API_KEY=gsk_your-groq-key\nMISTRAL_API_KEY=your-mistral-key\nMOONSHOT_API_KEY=sk-your-moonshot-key\n```\n\n### Quick Configuration Template\n\n```bash\n# Copy environment variable template\ncp env.example .env\n\n# Edit configuration file\nnano .env\n```\n\n**🔒 Security Tips**: \n- Never commit API keys to Git repositories\n- Use environment variables for sensitive information\n- Rotate API keys regularly\n\n---\n\n## 🌐 Web Interface\n\nAfter starting the service, you can access the management interface in two ways:\n\n### 🚀 Method 1: Integrated Web Interface (Default)\n```bash\nclaude-llm-gateway start --port 8765\n# Visit: http://localhost:8765\n```\n\n### 🌟 Method 2: Dedicated Web Server (Recommended)\n```bash\n# Start Gateway API service\nclaude-llm-gateway start --port 8765 --daemon\n\n# Start dedicated Web UI server  \ncd claude-llm-gateway && npm run web\n# Visit: http://localhost:9000\n\n# Or start both services together\nnpm run both\n```\n\n**🎯 Benefits of Method 2:**\n- **Better Performance**: Web UI and API separated\n- **Independent Scaling**: Scale web and API separately  \n- **Enhanced Security**: API and UI can run on different networks\n- **Development Friendly**: Hot reload for UI development\n\n### Interface Overview\n\nThe Claude LLM Gateway features a comprehensive web management interface with multiple functional tabs:\n\n#### 📊 Main Dashboard\n![Claude LLM Gateway Dashboard](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/dashboard.png)\n\nThe main dashboard provides real-time status overview with provider health monitoring and system statistics.\n\n#### 🔝 Top Status Bar\n- **Language Toggle**: One-click Chinese/English switching\n- **Refresh Status**: Manually refresh all provider statuses\n- **Last Updated**: Display data update timestamp\n\n#### 📊 Status Overview\n- **Healthy Providers**: Number of currently available models (shown as 2 in the image)\n- **Total Providers**: Total supported provider count (36)\n- **Available Models**: Total number of available models (78)\n- **Total Requests**: Request statistics\n\n#### 📝 Main Tabs\n\n**1. Provider Management** (as shown in image)\n- View real-time status of all LLM providers\n- Green toggle indicates enabled, red dot indicates unhealthy\n- Display request count, response time, cost information for each provider\n- Supported model lists (e.g., OpenAI's gpt-4, gpt-3.5-turbo, etc.)\n\n**2. Configuration** \n![Configuration Interface](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/configs.png)\n\n- Environment variable configuration with secure API key management\n- Real-time gateway settings adjustment\n- Provider-specific configuration options\n\n**3. API Endpoints**\n![API Endpoints](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/api-endpoints.png)\n\n- Complete API documentation with live examples\n- Claude-compatible endpoints with usage instructions\n- Token management API references\n- Health check and provider status interfaces\n\n**4. Intelligent Selection**\n![Model Selection Interface](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/modelsection.png)\n\n- Comprehensive model capability matrix with detailed statistics\n- Performance analytics and optimization insights\n- Task-based intelligent selection strategies\n- Real-time model comparison and recommendations\n\n**5. Real-time Logs**\n![Real-time Logs](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/realtime-logs.png)\n\n- Live system activity monitoring with auto-scroll\n- Detailed request/response logging\n- Provider health status updates\n- System diagnostic information\n\n#### 🎛️ Provider Card Operations\nEach provider card contains:\n- **Status Indicator**: Healthy/unhealthy status\n- **Enable Toggle**: One-click enable/disable provider\n- **Test Button**: Test provider connection\n- **Configure Button**: Quick jump to configuration page\n- **Error Messages**: Display specific error information (e.g., \"API key not configured\")\n\n---\n\n## 🛠️ CLI Commands\n\n### Basic Commands\n\n```bash\n# Start service\nclaude-llm-gateway start [options]\n\n# Run in background (daemon mode)\nclaude-llm-gateway start --daemon\n\n# Stop service\nclaude-llm-gateway stop\n\n# Check status\nclaude-llm-gateway status\n\n# Show help\nclaude-llm-gateway --help\n```\n\n### Web Server Commands\n\n```bash\n# Start dedicated web server (port 9000)\ncd claude-llm-gateway && npm run web\n\n# Start web server in development mode\nnpm run web:dev\n\n# Start both API gateway and web server\nnpm run both\n```\n\n### Startup Options\n\n```bash\n# Specify port (default: 8765)\nclaude-llm-gateway start --port 3000\n\n# Specify host (default: localhost)\nclaude-llm-gateway start --host 0.0.0.0\n\n# Run in background (daemon mode)\nclaude-llm-gateway start --daemon\n\n# Combined usage\nclaude-llm-gateway start --port 8765 --host 0.0.0.0 --daemon\n```\n\n### Status Management\n\n```bash\n# View detailed status\nclaude-llm-gateway status\n\n# Check all provider statuses\ncurl http://localhost:8765/providers\n\n# Health check\ncurl http://localhost:8765/health\n```\n\n---\n\n## 🔧 API Usage\n\n### Claude-Compatible Interface\n\nClaude LLM Gateway provides fully compatible Claude API interfaces:\n\n```bash\n# Claude Messages API\ncurl -X POST http://localhost:8765/v1/messages \\\n  -H \"Content-Type: application/json\" \\\n  -H \"x-api-key: your-api-key\" \\\n  -d '{\n    \"model\": \"claude-3-sonnet-20240229\",\n    \"max_tokens\": 1024,\n    \"messages\": [\n      {\n        \"role\": \"user\", \n        \"content\": \"Hello, world!\"\n      }\n    ]\n  }'\n```\n\n### OpenAI-Compatible Interface\n\nAlso supports OpenAI ChatGPT interface format:\n\n```bash\n# OpenAI Chat Completions API\ncurl -X POST http://localhost:8765/v1/chat/completions \\\n  -H \"Content-Type: application/json\" \\\n  -H \"Authorization: Bearer your-api-key\" \\\n  -d '{\n    \"model\": \"gpt-4\",\n    \"messages\": [\n      {\n        \"role\": \"user\",\n        \"content\": \"Hello, world!\"\n      }\n    ]\n  }'\n```\n\n### Intelligent Routing\n\nWhen no model is specified, the gateway automatically selects the optimal model:\n\n```bash\ncurl -X POST http://localhost:8765/v1/messages \\\n  -H \"Content-Type: application/json\" \\\n  -d '{\n    \"max_tokens\": 1024,\n    \"messages\": [\n      {\n        \"role\": \"user\",\n        \"content\": \"Write a Python function to calculate Fibonacci numbers\"\n      }\n    ]\n  }'\n```\n\n### Node.js SDK Usage\n\n```javascript\nconst { ClaudeLLMGateway } = require('claude-llm-gateway');\n\n// Create gateway instance\nconst gateway = new ClaudeLLMGateway();\n\n// Start service\nawait gateway.start(8765);\n\n// Use built-in client\nconst response = await gateway.chat({\n  messages: [\n    { role: 'user', content: 'Hello!' }\n  ]\n});\n\nconsole.log(response.content);\n```\n\n---\n\n## 🌍 Supported Models\n\n### OpenAI Models\n- **GPT-4o** (gpt-4o) - Latest multimodal model\n- **GPT-4 Turbo** (gpt-4-turbo) - High-performance version\n- **GPT-4** (gpt-4) - Classic version\n- **GPT-4o Mini** (gpt-4o-mini) - Lightweight version\n- **GPT-3.5 Turbo** (gpt-3.5-turbo) - Fast response\n\n### Anthropic Models  \n- **Claude 3 Opus** (claude-3-opus) - Strongest reasoning capabilities\n- **Claude 3.5 Sonnet** (claude-3.5-sonnet) - Balanced performance\n- **Claude 3 Sonnet** (claude-3-sonnet) - Standard version\n- **Claude 3 Haiku** (claude-3-haiku) - Fast response\n\n### Google Models\n- **Gemini Ultra** (gemini-ultra) - Flagship model\n- **Gemini 1.5 Pro** (gemini-1.5-pro) - Long context\n- **Gemini Pro** (gemini-pro) - Standard version\n\n### Meta Models\n- **Llama 3.1 405B** (llama-3.1-405b) - Ultra-large model\n- **Llama 3.1 70B** (llama-3.1-70b) - High performance\n- **Llama 3.1 8B** (llama-3.1-8b) - Lightweight\n\n### Chinese-Optimized Models\n- **DeepSeek V3** (deepseek-v3) - Chinese programming\n- **DeepSeek Chat** (deepseek-chat) - Chinese conversation\n- **DeepSeek Coder** (deepseek-coder) - Code generation\n- **Qianwen Max** (qianwen-max) - Alibaba's large model\n- **GLM-4** (zhipu-glm-4) - Zhipu AI\n- **Moonshot V1** (moonshot-v1) - Dark Side of the Moon\n\n### Other Models\n- **Mistral Large** (mistral-large) - European AI\n- **Groq Models** - Ultra-fast inference\n- **Cohere Command R+** - Enterprise-grade\n- **Grok-1** (grok-1) - xAI humor model\n\n**Total**: 30+ model providers, 78+ available models\n\n---\n\n## 📊 Intelligent Selection\n\n### Task Type Detection\n\nThe gateway can automatically detect task types and select optimal models:\n\n#### 🔧 Programming Tasks\n```\nInput: \"Write a Python function to calculate prime numbers\"\nRecommended: DeepSeek Coder, GPT-4, Claude 3.5 Sonnet\nReason: Strong code generation capabilities, accurate syntax\n```\n\n#### 📝 Creative Tasks  \n```\nInput: \"Write an article about AI\"\nRecommended: Claude 3 Opus, GPT-4, Gemini Pro\nReason: Strong creative abilities, natural writing style\n```\n\n#### 🔍 Analysis Tasks\n```\nInput: \"Analyze the trends in this data report\"\nRecommended: Claude 3 Opus, GPT-4 Turbo, Gemini Ultra  \nReason: Strong logical reasoning, deep analysis\n```\n\n#### 💬 Conversation Tasks\n```\nInput: \"Hello, please introduce yourself\"\nRecommended: GPT-3.5 Turbo, Claude 3 Haiku, DeepSeek Chat\nReason: Fast response, low cost, natural conversation\n```\n\n### Cost Optimization Strategies\n\n- **Intelligent Downgrading**: Automatically use lower-cost models for simple tasks\n- **Parallel Requests**: Send requests to multiple providers simultaneously, choose fastest response\n- **Failover**: Automatically switch to backup models when primary models are unavailable\n- **Load Balancing**: Distribute requests based on provider response time and success rate\n\n## 🔐 Security Configuration\n\n### API Key Management\n\n```bash\n# Recommended: Use environment variables\nexport OPENAI_API_KEY=\"sk-your-key\"\n\n# Avoid: Hard-coding in code\n# const apiKey = \"sk-your-key\"  ❌\n```\n\n### Access Control\n\n```bash\n# Restrict access sources\nclaude-llm-gateway start --host 127.0.0.1  # Local access only\n\n# Production environment\nclaude-llm-gateway start --host 0.0.0.0    # Allow external access\n```\n\n### Security Measures\n\n- ✅ **CORS Configuration**: Prevent cross-origin attacks\n- ✅ **Rate Limiting**: Prevent API abuse  \n- ✅ **Helmet Security Headers**: Enhanced security\n- ✅ **Input Validation**: Prevent injection attacks\n- ✅ **Error Handling**: No sensitive information leakage\n\n---\n\n## 📝 Troubleshooting\n\n### Common Issues\n\n**Q: Port already in use error on startup**\n```bash\n# Check port usage\nlsof -i :8765\n\n# Use different port\nclaude-llm-gateway start --port 8766\n```\n\n**Q: Invalid API key error**\n```bash\n# Check environment variables\necho $OPENAI_API_KEY\n\n# Reset key\nexport OPENAI_API_KEY=\"sk-your-new-key\"\n```\n\n**Q: Cannot connect to a provider**\n```bash\n# Check network connection\ncurl -I https://api.openai.com\n\n# View detailed errors\nclaude-llm-gateway start --verbose\n```\n\n**Q: Intelligent selection not working**\n```bash\n# View model capability configuration\ncurl http://localhost:8765/model-stats\n\n# Check provider status\ncurl http://localhost:8765/providers\n```\n\n### Debug Mode\n\n```bash\n# Enable detailed logging\nexport LOG_LEVEL=debug\nclaude-llm-gateway start\n\n# View real-time logs\ntail -f logs/gateway.log\n```\n\n### Performance Optimization\n\n```bash\n# Increase concurrent connections\nexport MAX_CONNECTIONS=100\n\n# Enable caching\nexport ENABLE_CACHE=true\n\n# Set timeout\nexport REQUEST_TIMEOUT=30000\n```\n\n---\n\n## 🤝 Contributing\n\nWelcome to contribute code! Please follow these steps:\n\n1. **Fork the repository** \n2. **Create feature branch**: `git checkout -b feature/amazing-feature`\n3. **Commit changes**: `git commit -m 'Add amazing feature'`\n4. **Push branch**: `git push origin feature/amazing-feature`\n5. **Create Pull Request**\n\n### Development Environment\n\n```bash\n# Clone repository\ngit clone https://github.com/chenxingqiang/claude-code-jailbreak.git\ncd claude-code-jailbreak/claude-llm-gateway\n\n# Install dependencies\nnpm install\n\n# Run tests\nnpm test\n\n# Start development mode\nnpm run dev\n```\n\n---\n\n## 📄 License\n\nThis project is licensed under the [MIT License](LICENSE)\n\n---\n\n## 🌟 Star History\n\nIf this project helps you, please give us a ⭐!\n\n[![Star History Chart](https://api.star-history.com/svg?repos=chenxingqiang/claude-code-jailbreak&type=Date)](https://star-history.com/#chenxingqiang/claude-code-jailbreak&Date)\n\n---\n\n## 📞 Support & Contact\n\n- 📁 **GitHub**: [claude-code-jailbreak](https://github.com/chenxingqiang/claude-code-jailbreak)\n- 📦 **NPM**: [claude-llm-gateway](https://www.npmjs.com/package/claude-llm-gateway)\n- 🐛 **Issues**: [GitHub Issues](https://github.com/chenxingqiang/claude-code-jailbreak/issues)\n- 💬 **Discussions**: [GitHub Discussions](https://github.com/chenxingqiang/claude-code-jailbreak/discussions)\n\n---\n\n**🎯 Making AI model selection simple and intelligent!**\n"}, "eac187fa41b08e9ea683e17f2d6773ccd39b8fcc": {"t": 1785765133.8500607, "n": 15934, "h": "c3b5f9adb0809da6", "c": "# 🧠 Claude LLM Gateway - Intelligent Multi-LLM API Gateway\n\n**Intelligent Multi-Model API Gateway - Unified Interface for 36+ Large Language Models**\n\n[![NPM Version](https://img.shields.io/npm/v/claude-llm-gateway)](https://www.npmjs.com/package/claude-llm-gateway)\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)\n[![Node.js](https://img.shields.io/badge/Node.js-16%2B-green)](https://nodejs.org/)\n\nClaude LLM Gateway is an intelligent API gateway that provides a unified multi-model LLM interface for Claude Code and other applications. It features intelligent model selection, cost optimization, load balancing, and failover capabilities.\n\n![Claude LLM Gateway Dashboard](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/dashboard.png)\n\n## ✨ Key Features\n\n- 🧠 **Intelligent Model Selection** - Auto-select optimal models based on task types\n- 🌍 **36+ Model Support** - OpenAI, Anthropic, Google, Meta, DeepSeek, and more\n- 🔄 **Bilingual Interface** - Complete Chinese/English internationalization\n- 💰 **Cost Optimization** - Smart routing to reduce API call costs\n- 📊 **Real-time Monitoring** - Web interface for all provider status monitoring\n- 🛡️ **Failover Protection** - Automatic switching to available backup models\n- 🔐 **Secure Gateway** - API key management and access control\n\n---\n\n## 📋 Table of Contents\n\n1. [🚀 Quick Start](#-quick-start)\n2. [📦 Installation](#-installation)\n3. [⚙️ Configuration](#-configuration)\n4. [🌐 Web Interface](#-web-interface)\n5. [🛠️ CLI Commands](#-cli-commands)\n6. [🔧 API Usage](#-api-usage)\n7. [🌍 Supported Models](#-supported-models)\n8. [📊 Intelligent Selection](#-intelligent-selection)\n9. [🔐 Security Configuration](#-security-configuration)\n10. [📝 Troubleshooting](#-troubleshooting)\n\n---\n\n## 🚀 Quick Start\n\n### System Requirements\n\n- **Node.js** v16+ - [Download here](https://nodejs.org/)\n- **npm** v7+ (comes with Node.js)\n- **Supported OS**: macOS, Linux, Windows\n\n### One-Click Installation\n\n```bash\n# Install globally via NPM\nnpm install -g claude-llm-gateway\n\n# Start immediately\nnpx claude-llm-gateway start --port 8765\n```\n\n🎉 **Done!** Open your browser and visit http://localhost:8765\n\n---\n\n## 📦 Installation\n\n### Method 1: Global Installation (Recommended)\n```bash\n# Install globally\nnpm install -g claude-llm-gateway\n\n# Start service\nclaude-llm-gateway start --port 8765\n\n# Run in background\nclaude-llm-gateway start --port 8765 --daemon\n```\n\n### Method 2: Local Project Installation\n```bash\n# Add to project dependencies\nnpm install claude-llm-gateway\n\n# Run with npx\nnpx claude-llm-gateway start --port 8765\n```\n\n### Method 3: Build from Source\n```bash\n# Clone repository\ngit clone https://github.com/chenxingqiang/claude-code-jailbreak.git\ncd claude-code-jailbreak/claude-llm-gateway\n\n# Install dependencies\nnpm install\n\n# Start development mode\nnpm run dev\n```\n\n## ⚙️ Configuration\n\n### Environment Variable Setup\n\nCreate a `.env` file or set environment variables:\n\n```bash\n# OpenAI\nOPENAI_API_KEY=sk-your-openai-key\n\n# Anthropic Claude\nANTHROPIC_API_KEY=sk-ant-your-anthropic-key\n\n# Google Gemini\nGOOGLE_API_KEY=your-google-api-key\n\n# DeepSeek (Recommended, large free quota)\nDEEPSEEK_API_KEY=sk-your-deepseek-key\n\n# Other optional configurations\nGROQ_API_KEY=gsk_your-groq-key\nMISTRAL_API_KEY=your-mistral-key\nMOONSHOT_API_KEY=sk-your-moonshot-key\n```\n\n### Quick Configuration Template\n\n```bash\n# Copy environment variable template\ncp env.example .env\n\n# Edit configuration file\nnano .env\n```\n\n**🔒 Security Tips**: \n- Never commit API keys to Git repositories\n- Use environment variables for sensitive information\n- Rotate API keys regularly\n\n---\n\n## 🌐 Web Interface\n\nAfter starting the service, you can access the management interface in two ways:\n\n### 🚀 Method 1: Integrated Web Interface (Default)\n```bash\nclaude-llm-gateway start --port 8765\n# Visit: http://localhost:8765\n```\n\n### 🌟 Method 2: Dedicated Web Server (Recommended)\n```bash\n# Start Gateway API service\nclaude-llm-gateway start --port 8765 --daemon\n\n# Start dedicated Web UI server  \ncd claude-llm-gateway && npm run web\n# Visit: http://localhost:9000\n\n# Or start both services together\nnpm run both\n```\n\n**🎯 Benefits of Method 2:**\n- **Better Performance**: Web UI and API separated\n- **Independent Scaling**: Scale web and API separately  \n- **Enhanced Security**: API and UI can run on different networks\n- **Development Friendly**: Hot reload for UI development\n\n### Interface Overview\n\nThe Claude LLM Gateway features a comprehensive web management interface with multiple functional tabs:\n\n#### 📊 Main Dashboard\n![Claude LLM Gateway Dashboard](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/dashboard.png)\n\nThe main dashboard provides real-time status overview with provider health monitoring and system statistics.\n\n#### 🔝 Top Status Bar\n- **Language Toggle**: One-click Chinese/English switching\n- **Refresh Status**: Manually refresh all provider statuses\n- **Last Updated**: Display data update timestamp\n\n#### 📊 Status Overview\n- **Healthy Providers**: Number of currently available models (shown as 2 in the image)\n- **Total Providers**: Total supported provider count (36)\n- **Available Models**: Total number of available models (78)\n- **Total Requests**: Request statistics\n\n#### 📝 Main Tabs\n\n**1. Provider Management** (as shown in image)\n- View real-time status of all LLM providers\n- Green toggle indicates enabled, red dot indicates unhealthy\n- Display request count, response time, cost information for each provider\n- Supported model lists (e.g., OpenAI's gpt-4, gpt-3.5-turbo, etc.)\n\n**2. Configuration** \n![Configuration Interface](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/config-api-key.png)\n\n- Environment variable configuration with secure API key management\n- Real-time gateway settings adjustment\n- Provider-specific configuration options\n\n**3. API Endpoints**\n![API Endpoints](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/api-endpoints.png)\n\n- Complete API documentation with live examples\n- Claude-compatible endpoints with usage instructions\n- Token management API references\n- Health check and provider status interfaces\n\n**4. Intelligent Selection**\n![Model Selection Interface](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/modelsection.png)\n\n- Comprehensive model capability matrix with detailed statistics\n- Performance analytics and optimization insights\n- Task-based intelligent selection strategies\n- Real-time model comparison and recommendations\n\n**5. Real-time Logs**\n![Real-time Logs](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/realtime-logs.png)\n\n- Live system activity monitoring with auto-scroll\n- Detailed request/response logging\n- Provider health status updates\n- System diagnostic information\n\n#### 🎛️ Provider Card Operations\nEach provider card contains:\n- **Status Indicator**: Healthy/unhealthy status\n- **Enable Toggle**: One-click enable/disable provider\n- **Test Button**: Test provider connection\n- **Configure Button**: Quick jump to configuration page\n- **Error Messages**: Display specific error information (e.g., \"API key not configured\")\n\n---\n\n## 🛠️ CLI Commands\n\n### Basic Commands\n\n```bash\n# Start service\nclaude-llm-gateway start [options]\n\n# Run in background (daemon mode)\nclaude-llm-gateway start --daemon\n\n# Stop service\nclaude-llm-gateway stop\n\n# Check status\nclaude-llm-gateway status\n\n# Show help\nclaude-llm-gateway --help\n```\n\n### Web Server Commands\n\n```bash\n# Start dedicated web server (port 9000)\ncd claude-llm-gateway && npm run web\n\n# Start web server in development mode\nnpm run web:dev\n\n# Start both API gateway and web server\nnpm run both\n```\n\n### Startup Options\n\n```bash\n# Specify port (default: 8765)\nclaude-llm-gateway start --port 3000\n\n# Specify host (default: localhost)\nclaude-llm-gateway start --host 0.0.0.0\n\n# Run in background (daemon mode)\nclaude-llm-gateway start --daemon\n\n# Combined usage\nclaude-llm-gateway start --port 8765 --host 0.0.0.0 --daemon\n```\n\n### Status Management\n\n```bash\n# View detailed status\nclaude-llm-gateway status\n\n# Check all provider statuses\ncurl http://localhost:8765/providers\n\n# Health check\ncurl http://localhost:8765/health\n```\n\n---\n\n## 🔧 API Usage\n\n### Claude-Compatible Interface\n\nClaude LLM Gateway provides fully compatible Claude API interfaces:\n\n```bash\n# Claude Messages API\ncurl -X POST http://localhost:8765/v1/messages \\\n  -H \"Content-Type: application/json\" \\\n  -H \"x-api-key: your-api-key\" \\\n  -d '{\n    \"model\": \"claude-3-sonnet-20240229\",\n    \"max_tokens\": 1024,\n    \"messages\": [\n      {\n        \"role\": \"user\", \n        \"content\": \"Hello, world!\"\n      }\n    ]\n  }'\n```\n\n### OpenAI-Compatible Interface\n\nAlso supports OpenAI ChatGPT interface format:\n\n```bash\n# OpenAI Chat Completions API\ncurl -X POST http://localhost:8765/v1/chat/completions \\\n  -H \"Content-Type: application/json\" \\\n  -H \"Authorization: Bearer your-api-key\" \\\n  -d '{\n    \"model\": \"gpt-4\",\n    \"messages\": [\n      {\n        \"role\": \"user\",\n        \"content\": \"Hello, world!\"\n      }\n    ]\n  }'\n```\n\n### Intelligent Routing\n\nWhen no model is specified, the gateway automatically selects the optimal model:\n\n```bash\ncurl -X POST http://localhost:8765/v1/messages \\\n  -H \"Content-Type: application/json\" \\\n  -d '{\n    \"max_tokens\": 1024,\n    \"messages\": [\n      {\n        \"role\": \"user\",\n        \"content\": \"Write a Python function to calculate Fibonacci numbers\"\n      }\n    ]\n  }'\n```\n\n### Node.js SDK Usage\n\n```javascript\nconst { ClaudeLLMGateway } = require('claude-llm-gateway');\n\n// Create gateway instance\nconst gateway = new ClaudeLLMGateway();\n\n// Start service\nawait gateway.start(8765);\n\n// Use built-in client\nconst response = await gateway.chat({\n  messages: [\n    { role: 'user', content: 'Hello!' }\n  ]\n});\n\nconsole.log(response.content);\n```\n\n---\n\n## 🌍 Supported Models\n\n### OpenAI Models\n- **GPT-4o** (gpt-4o) - Latest multimodal model\n- **GPT-4 Turbo** (gpt-4-turbo) - High-performance version\n- **GPT-4** (gpt-4) - Classic version\n- **GPT-4o Mini** (gpt-4o-mini) - Lightweight version\n- **GPT-3.5 Turbo** (gpt-3.5-turbo) - Fast response\n\n### Anthropic Models  \n- **Claude 3 Opus** (claude-3-opus) - Strongest reasoning capabilities\n- **Claude 3.5 Sonnet** (claude-3.5-sonnet) - Balanced performance\n- **Claude 3 Sonnet** (claude-3-sonnet) - Standard version\n- **Claude 3 Haiku** (claude-3-haiku) - Fast response\n\n### Google Models\n- **Gemini Ultra** (gemini-ultra) - Flagship model\n- **Gemini 1.5 Pro** (gemini-1.5-pro) - Long context\n- **Gemini Pro** (gemini-pro) - Standard version\n\n### Meta Models\n- **Llama 3.1 405B** (llama-3.1-405b) - Ultra-large model\n- **Llama 3.1 70B** (llama-3.1-70b) - High performance\n- **Llama 3.1 8B** (llama-3.1-8b) - Lightweight\n\n### Chinese-Optimized Models\n- **DeepSeek V3** (deepseek-v3) - Chinese programming\n- **DeepSeek Chat** (deepseek-chat) - Chinese conversation\n- **DeepSeek Coder** (deepseek-coder) - Code generation\n- **Qianwen Max** (qianwen-max) - Alibaba's large model\n- **GLM-4** (zhipu-glm-4) - Zhipu AI\n- **Moonshot V1** (moonshot-v1) - Dark Side of the Moon\n\n### Other Models\n- **Mistral Large** (mistral-large) - European AI\n- **Groq Models** - Ultra-fast inference\n- **Cohere Command R+** - Enterprise-grade\n- **Grok-1** (grok-1) - xAI humor model\n\n**Total**: 30+ model providers, 78+ available models\n\n---\n\n## 📊 Intelligent Selection\n\n### Task Type Detection\n\nThe gateway can automatically detect task types and select optimal models:\n\n#### 🔧 Programming Tasks\n```\nInput: \"Write a Python function to calculate prime numbers\"\nRecommended: DeepSeek Coder, GPT-4, Claude 3.5 Sonnet\nReason: Strong code generation capabilities, accurate syntax\n```\n\n#### 📝 Creative Tasks  \n```\nInput: \"Write an article about AI\"\nRecommended: Claude 3 Opus, GPT-4, Gemini Pro\nReason: Strong creative abilities, natural writing style\n```\n\n#### 🔍 Analysis Tasks\n```\nInput: \"Analyze the trends in this data report\"\nRecommended: Claude 3 Opus, GPT-4 Turbo, Gemini Ultra  \nReason: Strong logical reasoning, deep analysis\n```\n\n#### 💬 Conversation Tasks\n```\nInput: \"Hello, please introduce yourself\"\nRecommended: GPT-3.5 Turbo, Claude 3 Haiku, DeepSeek Chat\nReason: Fast response, low cost, natural conversation\n```\n\n### Cost Optimization Strategies\n\n- **Intelligent Downgrading**: Automatically use lower-cost models for simple tasks\n- **Parallel Requests**: Send requests to multiple providers simultaneously, choose fastest response\n- **Failover**: Automatically switch to backup models when primary models are unavailable\n- **Load Balancing**: Distribute requests based on provider response time and success rate\n\n## 🔐 Security Configuration\n\n### API Key Management\n\n```bash\n# Recommended: Use environment variables\nexport OPENAI_API_KEY=\"sk-your-key\"\n\n# Avoid: Hard-coding in code\n# const apiKey = \"sk-your-key\"  ❌\n```\n\n### Access Control\n\n```bash\n# Restrict access sources\nclaude-llm-gateway start --host 127.0.0.1  # Local access only\n\n# Production environment\nclaude-llm-gateway start --host 0.0.0.0    # Allow external access\n```\n\n### Security Measures\n\n- ✅ **CORS Configuration**: Prevent cross-origin attacks\n- ✅ **Rate Limiting**: Prevent API abuse  \n- ✅ **Helmet Security Headers**: Enhanced security\n- ✅ **Input Validation**: Prevent injection attacks\n- ✅ **Error Handling**: No sensitive information leakage\n\n---\n\n## 📝 Troubleshooting\n\n### Common Issues\n\n**Q: Port already in use error on startup**\n```bash\n# Check port usage\nlsof -i :8765\n\n# Use different port\nclaude-llm-gateway start --port 8766\n```\n\n**Q: Invalid API key error**\n```bash\n# Check environment variables\necho $OPENAI_API_KEY\n\n# Reset key\nexport OPENAI_API_KEY=\"sk-your-new-key\"\n```\n\n**Q: Cannot connect to a provider**\n```bash\n# Check network connection\ncurl -I https://api.openai.com\n\n# View detailed errors\nclaude-llm-gateway start --verbose\n```\n\n**Q: Intelligent selection not working**\n```bash\n# View model capability configuration\ncurl http://localhost:8765/model-stats\n\n# Check provider status\ncurl http://localhost:8765/providers\n```\n\n### Debug Mode\n\n```bash\n# Enable detailed logging\nexport LOG_LEVEL=debug\nclaude-llm-gateway start\n\n# View real-time logs\ntail -f logs/gateway.log\n```\n\n### Performance Optimization\n\n```bash\n# Increase concurrent connections\nexport MAX_CONNECTIONS=100\n\n# Enable caching\nexport ENABLE_CACHE=true\n\n# Set timeout\nexport REQUEST_TIMEOUT=30000\n```\n\n---\n\n## 🤝 Contributing\n\nWelcome to contribute code! Please follow these steps:\n\n1. **Fork the repository** \n2. **Create feature branch**: `git checkout -b feature/amazing-feature`\n3. **Commit changes**: `git commit -m 'Add amazing feature'`\n4. **Push branch**: `git push origin feature/amazing-feature`\n5. **Create Pull Request**\n\n### Development Environment\n\n```bash\n# Clone repository\ngit clone https://github.com/chenxingqiang/claude-code-jailbreak.git\ncd claude-code-jailbreak/claude-llm-gateway\n\n# Install dependencies\nnpm install\n\n# Run tests\nnpm test\n\n# Start development mode\nnpm run dev\n```\n\n---\n\n## 📄 License\n\nThis project is licensed under the [MIT License](LICENSE)\n\n---\n\n## 🌟 Star History\n\nIf this project helps you, please give us a ⭐!\n\n[![Star History Chart](https://api.star-history.com/svg?repos=chenxingqiang/claude-code-jailbreak&type=Date)](https://star-history.com/#chenxingqiang/claude-code-jailbreak&Date)\n\n---\n\n## 📞 Support & Contact\n\n- 📁 **GitHub**: [claude-code-jailbreak](https://github.com/chenxingqiang/claude-code-jailbreak)\n- 📦 **NPM**: [claude-llm-gateway](https://www.npmjs.com/package/claude-llm-gateway)\n- 🐛 **Issues**: [GitHub Issues](https://github.com/chenxingqiang/claude-code-jailbreak/issues)\n- 💬 **Discussions**: [GitHub Discussions](https://github.com/chenxingqiang/claude-code-jailbreak/discussions)\n\n---\n\n**🎯 Making AI model selection simple and intelligent!**\n"}, "b9ba99806e0171441b0166141dd87c7799ebf8e6": {"t": 1785765136.4537392, "n": 15021, "h": "b680ba49e4e5d9f0", "c": "# 🧠 Claude LLM Gateway - Intelligent Multi-LLM API Gateway\n\n**Intelligent Multi-Model API Gateway - Unified Interface for 36+ Large Language Models**\n\n[![NPM Version](https://img.shields.io/npm/v/claude-llm-gateway)](https://www.npmjs.com/package/claude-llm-gateway)\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)\n[![Node.js](https://img.shields.io/badge/Node.js-16%2B-green)](https://nodejs.org/)\n\nClaude LLM Gateway is an intelligent API gateway that provides a unified multi-model LLM interface for Claude Code and other applications. It features intelligent model selection, cost optimization, load balancing, and failover capabilities.\n\n![Claude LLM Gateway Dashboard](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/dashboard.png)\n\n## ✨ Key Features\n\n- 🧠 **Intelligent Model Selection** - Auto-select optimal models based on task types\n- 🌍 **36+ Model Support** - OpenAI, Anthropic, Google, Meta, DeepSeek, and more\n- 🔄 **Bilingual Interface** - Complete Chinese/English internationalization\n- 💰 **Cost Optimization** - Smart routing to reduce API call costs\n- 📊 **Real-time Monitoring** - Web interface for all provider status monitoring\n- 🛡️ **Failover Protection** - Automatic switching to available backup models\n- 🔐 **Secure Gateway** - API key management and access control\n\n---\n\n## 📋 Table of Contents\n\n1. [🚀 Quick Start](#-quick-start)\n2. [📦 Installation](#-installation)\n3. [⚙️ Configuration](#-configuration)\n4. [🌐 Web Interface](#-web-interface)\n5. [🛠️ CLI Commands](#-cli-commands)\n6. [🔧 API Usage](#-api-usage)\n7. [🌍 Supported Models](#-supported-models)\n8. [📊 Intelligent Selection](#-intelligent-selection)\n9. [🔐 Security Configuration](#-security-configuration)\n10. [📝 Troubleshooting](#-troubleshooting)\n\n---\n\n## 🚀 Quick Start\n\n### System Requirements\n\n- **Node.js** v16+ - [Download here](https://nodejs.org/)\n- **npm** v7+ (comes with Node.js)\n- **Supported OS**: macOS, Linux, Windows\n\n### One-Click Installation\n\n```bash\n# Install globally via NPM\nnpm install -g claude-llm-gateway\n\n# Start immediately\nnpx claude-llm-gateway start --port 8765\n```\n\n🎉 **Done!** Open your browser and visit http://localhost:8765\n\n---\n\n## 📦 Installation\n\n### Method 1: Global Installation (Recommended)\n```bash\n# Install globally\nnpm install -g claude-llm-gateway\n\n# Start service\nclaude-llm-gateway start --port 8765\n\n# Run in background\nclaude-llm-gateway start --port 8765 --daemon\n```\n\n### Method 2: Local Project Installation\n```bash\n# Add to project dependencies\nnpm install claude-llm-gateway\n\n# Run with npx\nnpx claude-llm-gateway start --port 8765\n```\n\n### Method 3: Build from Source\n```bash\n# Clone repository\ngit clone https://github.com/chenxingqiang/claude-code-jailbreak.git\ncd claude-code-jailbreak/claude-llm-gateway\n\n# Install dependencies\nnpm install\n\n# Start development mode\nnpm run dev\n```\n\n## ⚙️ Configuration\n\n### Environment Variable Setup\n\nCreate a `.env` file or set environment variables:\n\n```bash\n# OpenAI\nOPENAI_API_KEY=sk-your-openai-key\n\n# Anthropic Claude\nANTHROPIC_API_KEY=sk-ant-your-anthropic-key\n\n# Google Gemini\nGOOGLE_API_KEY=your-google-api-key\n\n# DeepSeek (Recommended, large free quota)\nDEEPSEEK_API_KEY=sk-your-deepseek-key\n\n# Other optional configurations\nGROQ_API_KEY=gsk_your-groq-key\nMISTRAL_API_KEY=your-mistral-key\nMOONSHOT_API_KEY=sk-your-moonshot-key\n```\n\n### Quick Configuration Template\n\n```bash\n# Copy environment variable template\ncp env.example .env\n\n# Edit configuration file\nnano .env\n```\n\n**🔒 Security Tips**: \n- Never commit API keys to Git repositories\n- Use environment variables for sensitive information\n- Rotate API keys regularly\n\n---\n\n## 🌐 Web Interface\n\nAfter starting the service, visit http://localhost:8765 to access the management interface.\n\n### Interface Overview\n\nThe Claude LLM Gateway features a comprehensive web management interface with multiple functional tabs:\n\n#### 📊 Main Dashboard\n![Claude LLM Gateway Dashboard](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/dashboard.png)\n\nThe main dashboard provides real-time status overview with provider health monitoring and system statistics.\n\n#### 🔝 Top Status Bar\n- **Language Toggle**: One-click Chinese/English switching\n- **Refresh Status**: Manually refresh all provider statuses\n- **Last Updated**: Display data update timestamp\n\n#### 📊 Status Overview\n- **Healthy Providers**: Number of currently available models (shown as 2 in the image)\n- **Total Providers**: Total supported provider count (36)\n- **Available Models**: Total number of available models (78)\n- **Total Requests**: Request statistics\n\n#### 📝 Main Tabs\n\n**1. Provider Management** (as shown in image)\n- View real-time status of all LLM providers\n- Green toggle indicates enabled, red dot indicates unhealthy\n- Display request count, response time, cost information for each provider\n- Supported model lists (e.g., OpenAI's gpt-4, gpt-3.5-turbo, etc.)\n\n**2. Configuration** \n![Configuration Interface](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/cofig-api-key.png)\n\n- Environment variable configuration with secure API key management\n- Real-time gateway settings adjustment\n- Provider-specific configuration options\n\n**3. API Endpoints**\n![API Endpoints](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/api-endpoints.png)\n\n- Complete API documentation with live examples\n- Claude-compatible endpoints with usage instructions\n- Token management API references\n- Health check and provider status interfaces\n\n**4. Intelligent Selection**\n![Model Selection Interface](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/modelsection.png)\n\n- Comprehensive model capability matrix with detailed statistics\n- Performance analytics and optimization insights\n- Task-based intelligent selection strategies\n- Real-time model comparison and recommendations\n\n**5. Real-time Logs**\n![Real-time Logs](https://raw.githubusercontent.com/chenxingqiang/claude-code-jailbreak/main/asset/realtime-logs.png)\n\n- Live system activity monitoring with auto-scroll\n- Detailed request/response logging\n- Provider health status updates\n- System diagnostic information\n\n#### 🎛️ Provider Card Operations\nEach provider card contains:\n- **Status Indicator**: Healthy/unhealthy status\n- **Enable Toggle**: One-click enable/disable provider\n- **Test Button**: Test provider connection\n- **Configure Button**: Quick jump to configuration page\n- **Error Messages**: Display specific error information (e.g., \"API key not configured\")\n\n---\n\n## 🛠️ CLI Commands\n\n### Basic Commands\n\n```bash\n# Start service\nclaude-llm-gateway start [options]\n\n# Run in background\nclaude-llm-gateway start --daemon\n\n# Stop service\nclaude-llm-gateway stop\n\n# Check status\nclaude-llm-gateway status\n\n# Show help\nclaude-llm-gateway --help\n```\n\n### Startup Options\n\n```bash\n# Specify port (default: 8765)\nclaude-llm-gateway start --port 3000\n\n# Specify host (default: localhost)\nclaude-llm-gateway start --host 0.0.0.0\n\n# Run in background (daemon mode)\nclaude-llm-gateway start --daemon\n\n# Combined usage\nclaude-llm-gateway start --port 8765 --host 0.0.0.0 --daemon\n```\n\n### Status Management\n\n```bash\n# View detailed status\nclaude-llm-gateway status\n\n# Check all provider statuses\ncurl http://localhost:8765/providers\n\n# Health check\ncurl http://localhost:8765/health\n```\n\n---\n\n## 🔧 API Usage\n\n### Claude-Compatible Interface\n\nClaude LLM Gateway provides fully compatible Claude API interfaces:\n\n```bash\n# Claude Messages API\ncurl -X POST http://localhost:8765/v1/messages \\\n  -H \"Content-Type: application/json\" \\\n  -H \"x-api-key: your-api-key\" \\\n  -d '{\n    \"model\": \"claude-3-sonnet-20240229\",\n    \"max_tokens\": 1024,\n    \"messages\": [\n      {\n        \"role\": \"user\", \n        \"content\": \"Hello, world!\"\n      }\n    ]\n  }'\n```\n\n### OpenAI-Compatible Interface\n\nAlso supports OpenAI ChatGPT interface format:\n\n```bash\n# OpenAI Chat Completions API\ncurl -X POST http://localhost:8765/v1/chat/completions \\\n  -H \"Content-Type: application/json\" \\\n  -H \"Authorization: Bearer your-api-key\" \\\n  -d '{\n    \"model\": \"gpt-4\",\n    \"messages\": [\n      {\n        \"role\": \"user\",\n        \"content\": \"Hello, world!\"\n      }\n    ]\n  }'\n```\n\n### Intelligent Routing\n\nWhen no model is specified, the gateway automatically selects the optimal model:\n\n```bash\ncurl -X POST http://localhost:8765/v1/messages \\\n  -H \"Content-Type: application/json\" \\\n  -d '{\n    \"max_tokens\": 1024,\n    \"messages\": [\n      {\n        \"role\": \"user\",\n        \"content\": \"Write a Python function to calculate Fibonacci numbers\"\n      }\n    ]\n  }'\n```\n\n### Node.js SDK Usage\n\n```javascript\nconst { ClaudeLLMGateway } = require('claude-llm-gateway');\n\n// Create gateway instance\nconst gateway = new ClaudeLLMGateway();\n\n// Start service\nawait gateway.start(8765);\n\n// Use built-in client\nconst response = await gateway.chat({\n  messages: [\n    { role: 'user', content: 'Hello!' }\n  ]\n});\n\nconsole.log(response.content);\n```\n\n---\n\n## 🌍 Supported Models\n\n### OpenAI Models\n- **GPT-4o** (gpt-4o) - Latest multimodal model\n- **GPT-4 Turbo** (gpt-4-turbo) - High-performance version\n- **GPT-4** (gpt-4) - Classic version\n- **GPT-4o Mini** (gpt-4o-mini) - Lightweight version\n- **GPT-3.5 Turbo** (gpt-3.5-turbo) - Fast response\n\n### Anthropic Models  \n- **Claude 3 Opus** (claude-3-opus) - Strongest reasoning capabilities\n- **Claude 3.5 Sonnet** (claude-3.5-sonnet) - Balanced performance\n- **Claude 3 Sonnet** (claude-3-sonnet) - Standard version\n- **Claude 3 Haiku** (claude-3-haiku) - Fast response\n\n### Google Models\n- **Gemini Ultra** (gemini-ultra) - Flagship model\n- **Gemini 1.5 Pro** (gemini-1.5-pro) - Long context\n- **Gemini Pro** (gemini-pro) - Standard version\n\n### Meta Models\n- **Llama 3.1 405B** (llama-3.1-405b) - Ultra-large model\n- **Llama 3.1 70B** (llama-3.1-70b) - High performance\n- **Llama 3.1 8B** (llama-3.1-8b) - Lightweight\n\n### Chinese-Optimized Models\n- **DeepSeek V3** (deepseek-v3) - Chinese programming\n- **DeepSeek Chat** (deepseek-chat) - Chinese conversation\n- **DeepSeek Coder** (deepseek-coder) - Code generation\n- **Qianwen Max** (qianwen-max) - Alibaba's large model\n- **GLM-4** (zhipu-glm-4) - Zhipu AI\n- **Moonshot V1** (moonshot-v1) - Dark Side of the Moon\n\n### Other Models\n- **Mistral Large** (mistral-large) - European AI\n- **Groq Models** - Ultra-fast inference\n- **Cohere Command R+** - Enterprise-grade\n- **Grok-1** (grok-1) - xAI humor model\n\n**Total**: 30+ model providers, 78+ available models\n\n---\n\n## 📊 Intelligent Selection\n\n### Task Type Detection\n\nThe gateway can automatically detect task types and select optimal models:\n\n#### 🔧 Programming Tasks\n```\nInput: \"Write a Python function to calculate prime numbers\"\nRecommended: DeepSeek Coder, GPT-4, Claude 3.5 Sonnet\nReason: Strong code generation capabilities, accurate syntax\n```\n\n#### 📝 Creative Tasks  \n```\nInput: \"Write an article about AI\"\nRecommended: Claude 3 Opus, GPT-4, Gemini Pro\nReason: Strong creative abilities, natural writing style\n```\n\n#### 🔍 Analysis Tasks\n```\nInput: \"Analyze the trends in this data report\"\nRecommended: Claude 3 Opus, GPT-4 Turbo, Gemini Ultra  \nReason: Strong logical reasoning, deep analysis\n```\n\n#### 💬 Conversation Tasks\n```\nInput: \"Hello, please introduce yourself\"\nRecommended: GPT-3.5 Turbo, Claude 3 Haiku, DeepSeek Chat\nReason: Fast response, low cost, natural conversation\n```\n\n### Cost Optimization Strategies\n\n- **Intelligent Downgrading**: Automatically use lower-cost models for simple tasks\n- **Parallel Requests**: Send requests to multiple providers simultaneously, choose fastest response\n- **Failover**: Automatically switch to backup models when primary models are unavailable\n- **Load Balancing**: Distribute requests based on provider response time and success rate\n\n## 🔐 Security Configuration\n\n### API Key Management\n\n```bash\n# Recommended: Use environment variables\nexport OPENAI_API_KEY=\"sk-your-key\"\n\n# Avoid: Hard-coding in code\n# const apiKey = \"sk-your-key\"  ❌\n```\n\n### Access Control\n\n```bash\n# Restrict access sources\nclaude-llm-gateway start --host 127.0.0.1  # Local access only\n\n# Production environment\nclaude-llm-gateway start --host 0.0.0.0    # Allow external access\n```\n\n### Security Measures\n\n- ✅ **CORS Configuration**: Prevent cross-origin attacks\n- ✅ **Rate Limiting**: Prevent API abuse  \n- ✅ **Helmet Security Headers**: Enhanced security\n- ✅ **Input Validation**: Prevent injection attacks\n- ✅ **Error Handling**: No sensitive information leakage\n\n---\n\n## 📝 Troubleshooting\n\n### Common Issues\n\n**Q: Port already in use error on startup**\n```bash\n# Check port usage\nlsof -i :8765\n\n# Use different port\nclaude-llm-gateway start --port 8766\n```\n\n**Q: Invalid API key error**\n```bash\n# Check environment variables\necho $OPENAI_API_KEY\n\n# Reset key\nexport OPENAI_API_KEY=\"sk-your-new-key\"\n```\n\n**Q: Cannot connect to a provider**\n```bash\n# Check network connection\ncurl -I https://api.openai.com\n\n# View detailed errors\nclaude-llm-gateway start --verbose\n```\n\n**Q: Intelligent selection not working**\n```bash\n# View model capability configuration\ncurl http://localhost:8765/model-stats\n\n# Check provider status\ncurl http://localhost:8765/providers\n```\n\n### Debug Mode\n\n```bash\n# Enable detailed logging\nexport LOG_LEVEL=debug\nclaude-llm-gateway start\n\n# View real-time logs\ntail -f logs/gateway.log\n```\n\n### Performance Optimization\n\n```bash\n# Increase concurrent connections\nexport MAX_CONNECTIONS=100\n\n# Enable caching\nexport ENABLE_CACHE=true\n\n# Set timeout\nexport REQUEST_TIMEOUT=30000\n```\n\n---\n\n## 🤝 Contributing\n\nWelcome to contribute code! Please follow these steps:\n\n1. **Fork the repository** \n2. **Create feature branch**: `git checkout -b feature/amazing-feature`\n3. **Commit changes**: `git commit -m 'Add amazing feature'`\n4. **Push branch**: `git push origin feature/amazing-feature`\n5. **Create Pull Request**\n\n### Development Environment\n\n```bash\n# Clone repository\ngit clone https://github.com/chenxingqiang/claude-code-jailbreak.git\ncd claude-code-jailbreak/claude-llm-gateway\n\n# Install dependencies\nnpm install\n\n# Run tests\nnpm test\n\n# Start development mode\nnpm run dev\n```\n\n---\n\n## 📄 License\n\nThis project is licensed under the [MIT License](LICENSE)\n\n---\n\n## 🌟 Star History\n\nIf this project helps you, please give us a ⭐!\n\n[![Star History Chart](https://api.star-history.com/svg?repos=chenxingqiang/claude-code-jailbreak&type=Date)](https://star-history.com/#chenxingqiang/claude-code-jailbreak&Date)\n\n---\n\n## 📞 Support & Contact\n\n- 📁 **GitHub**: [claude-code-jailbreak](https://github.com/chenxingqiang/claude-code-jailbreak)\n- 📦 **NPM**: [claude-llm-gateway](https://www.npmjs.com/package/claude-llm-gateway)\n- 🐛 **Issues**: [GitHub Issues](https://github.com/chenxingqiang/claude-code-jailbreak/issues)\n- 💬 **Discussions**: [GitHub Discussions](https://github.com/chenxingqiang/claude-code-jailbreak/discussions)\n\n---\n\n**🎯 Making AI model selection simple and intelligent!**\n"}, "5d835c4a869544133ba604665702cfbd40b3ed33": {"t": 1785765137.824092, "n": 9897, "h": "ea4a2b59e252d0ce", "c": "# 🚀 Claude Code Jailbreak Tutorial\n\n**Complete guide to bypass Claude CLI organization restrictions using Moonshot API**\n\nThis comprehensive tutorial shows you how to install Claude Code, bypass organization authentication, and use your own Moonshot API key instead.\n\n---\n\n## 📋 Table of Contents\n\n1. [Prerequisites](#prerequisites)\n2. [Step 1: Install Claude Code](#step-1-install-claude-code)\n3. [Step 2: Get Moonshot API Key](#step-2-get-moonshot-api-key)\n4. [Step 3: Jailbreak Setup](#step-3-jailbreak-setup)\n5. [Step 4: Create GitHub Repository](#step-4-create-github-repository)\n6. [Step 5: Testing & Verification](#step-5-testing--verification)\n7. [Usage Examples](#usage-examples)\n8. [Troubleshooting](#troubleshooting)\n\n---\n\n## Prerequisites\n\n- **macOS/Linux/Windows** with terminal access\n- **Node.js** (v16 or higher) - [Download here](https://nodejs.org/)\n- **Git** - [Download here](https://git-scm.com/)\n- **GitHub account** - [Sign up here](https://github.com/)\n\n---\n\n## Step 1: Install Claude Code\n\n### Option A: Direct Installation (Recommended)\n```bash\n# Install Claude Code globally\nnpm install -g @anthropic-ai/claude-code\n\n# Verify installation\nclaude --version\n```\n\n### Option B: Alternative Installation Methods\n```bash\n# Using curl (Linux/macOS)\ncurl -fsSL https://claude.ai/claude-code/install.sh | sh\n\n# Using PowerShell (Windows)\nirm https://claude.ai/claude-code/install.ps1 | iex\n```\n\n### Verify Installation\n```bash\n# Check if Claude is installed correctly\nwhich claude\nclaude --help\n```\n\n---\n\n## Step 2: Get Moonshot API Key\n\n### 2.1 Register at Moonshot AI\n1. Visit [Moonshot AI Console](https://platform.moonshot.cn/)\n2. Create an account or sign in\n3. Navigate to API Keys section\n4. Generate a new API key\n\n### 2.2 API Key Format\nYour key should look like: `sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx`\n\n**🔒 Security Note**: Never share your API key publicly!\n\n---\n\n## Step 3: Jailbreak Setup\n\n### 3.1 Create Project Directory\n```bash\n# Create and navigate to project directory\nmkdir claude-code-jailbreak\ncd claude-code-jailbreak\n\n# Initialize git repository\ngit init\n```\n\n### 3.2 Create Environment Configuration\n\nCreate the jailbreak environment file:\n\n```bash\ncat > claude-env.sh << 'EOF'\n#!/bin/bash\n# Claude CLI Jailbreak Configuration for Moonshot API\n# This bypasses organization authentication\n\n# Decode base64 encoded API key (security through obscurity)\nENCODED_KEY=\"c2stNTBmR2I2ZzlYN2w2N1ZOQzFKQ0VoMWtjclFJQmVnbEhyUHN6MDlSSTVmWGcwSVQx\"\nDECODED_KEY=$(echo $ENCODED_KEY | base64 -d)\n\n# Set environment variables for Claude CLI\nexport ANTHROPIC_API_KEY=\"$DECODED_KEY\"\nexport ANTHROPIC_BASE_URL=\"https://api.moonshot.cn/anthropic\"\n\n# Alternative variables for compatibility\nexport MOONSHOT_API_KEY=\"$DECODED_KEY\"\nexport MOONSHOT_BASE_URL=\"https://api.moonshot.cn/anthropic\"\nexport ANTHROPIC_AUTH_TOKEN=\"$DECODED_KEY\"\n\n# Success message\necho \"🔓 Claude CLI Jailbreak Activated!\"\necho \"🔑 API Key: ${ANTHROPIC_API_KEY:0:10}...\"\necho \"🌐 Endpoint: $ANTHROPIC_BASE_URL\"\necho \"✅ Ready to bypass organization restrictions!\"\necho \"\"\necho \"Usage: claude --print 'your prompt here'\"\nEOF\n\n# Make executable\nchmod +x claude-env.sh\n```\n\n### 3.3 Create Configuration File\n```bash\ncat > .claude-jailbreak-config << 'EOF'\n{\n  \"version\": \"1.0.0\",\n  \"provider\": \"moonshot\",\n  \"bypass_auth\": true,\n  \"api_endpoint\": \"https://api.moonshot.cn/anthropic\",\n  \"features\": {\n    \"no_organization_required\": true,\n    \"custom_api_key\": true,\n    \"bypass_restrictions\": true\n  }\n}\nEOF\n```\n\n---\n\n## Step 4: Create GitHub Repository\n\n### 4.1 Create Repository on GitHub\n```bash\n# Login to GitHub CLI (if not already)\ngh auth login\n\n# Create repository\ngh repo create claude-code-jailbreak --public --description \"Claude CLI jailbreak using Moonshot API to bypass organization restrictions\"\n\n# Add remote origin  \ngit remote add origin https://github.com/chenxingqiang/claude-code-jailbreak.git\n```\n\n### 4.2 Create Additional Files\n\n**Create .gitignore:**\n```bash\ncat > .gitignore << 'EOF'\n# Security - Never commit real API keys\n*.env\n.env.*\n**/secrets/**\n**/*secret*\n**/*key*\n\n# Logs\n*.log\nlogs/\n\n# OS Files\n.DS_Store\nThumbs.db\n\n# IDE\n.vscode/\n.idea/\n*.swp\n*.swo\n\n# Dependencies\nnode_modules/\nEOF\n```\n\n**Create LICENSE:**\n```bash\ncat > LICENSE << 'EOF'\nMIT License\n\nCopyright (c) 2024 Claude Code Jailbreak\n\nPermission is hereby granted, free of charge, to any person obtaining a copy\nof this software and associated documentation files (the \"Software\"), to deal\nin the Software without restriction, including without limitation the rights\nto use, copy, modify, merge, publish, distribute, sublicense, and/or sell\ncopies of the Software, and to permit persons to whom the Software is\nfurnished to do so, subject to the following conditions:\n\nThe above copyright notice and this permission notice shall be included in all\ncopies or substantial portions of the Software.\n\nTHE SOFTWARE IS PROVIDED \"AS IS\", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR\nIMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,\nFITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE\nAUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER\nLIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,\nOUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE\nSOFTWARE.\nEOF\n```\n\n### 4.3 Commit and Push\n```bash\n# Stage all files\ngit add .\n\n# Commit changes\ngit commit -m \"🚀 Initial commit: Claude Code jailbreak setup with Moonshot API\"\n\n# Push to GitHub\ngit push -u origin main\n```\n\n---\n\n## Step 5: Testing & Verification\n\n### 5.1 Test the Jailbreak\n```bash\n# Activate jailbreak environment\nsource claude-env.sh\n\n# Test basic functionality\nclaude --print \"Hello! Are you working without organization authentication?\"\n\n# Test with more complex prompt\nclaude --print \"Write a simple Python function to calculate fibonacci numbers\"\n```\n\n### 5.2 Verify Success Indicators\nYou should see:\n- ✅ No organization authentication errors\n- ✅ Successful API responses\n- ✅ Custom endpoint being used\n- ✅ API key properly loaded\n\n---\n\n## Usage Examples\n\n### Quick Start\n```bash\n# Every time you want to use Claude CLI:\nsource claude-env.sh\nclaude --print \"Your prompt here\"\n```\n\n### Interactive Mode\n```bash\nsource claude-env.sh\nclaude\n# Now you can chat interactively!\n```\n\n### Advanced Usage\n```bash\n# With specific model\nclaude --print --model sonnet \"Explain quantum computing\"\n\n# With output formatting\nclaude --print --output-format json \"List 5 programming languages\"\n\n# Continue conversation\nclaude --continue\n```\n\n### Permanent Setup (Optional)\nAdd to your `~/.zshrc` or `~/.bashrc`:\n```bash\n# Auto-load Claude jailbreak on terminal start\nif [ -f \"$HOME/claude-code-jailbreak/claude-env.sh\" ]; then\n    source \"$HOME/claude-code-jailbreak/claude-env.sh\"\nfi\n```\n\n---\n\n## 🔧 Troubleshooting\n\n### Common Issues\n\n**Error: \"This organization has been disabled\"**\n- ✅ **Solution**: Use this jailbreak! It bypasses organization auth entirely.\n\n**Error: \"API key not found\"**\n- Check if environment variables are set: `echo $ANTHROPIC_API_KEY`\n- Re-run: `source claude-env.sh`\n- Verify your Moonshot API key is valid\n\n**Error: \"Connection refused\"**\n- Check your internet connection\n- Verify Moonshot API endpoint is accessible\n- Ensure your API key has sufficient credits\n\n**Error: \"Invalid base64 encoding\"**\n- The encoded key might be corrupted\n- Replace with your actual API key in the script\n\n### Debug Mode\n```bash\n# Enable debug mode for detailed error information\nclaude --debug --print \"test prompt\"\n```\n\n### Manual Configuration\nIf the automated setup doesn't work, manually set:\n```bash\nexport ANTHROPIC_API_KEY=\"your-actual-moonshot-api-key\"\nexport ANTHROPIC_BASE_URL=\"https://api.moonshot.cn/anthropic\"\n```\n\n---\n\n## 🔒 Security Notes\n\n1. **API Key Encoding**: The tutorial uses base64 encoding for basic obfuscation\n2. **Never commit real keys**: Always use environment variables or encoded formats\n3. **Regular rotation**: Rotate your API keys periodically\n4. **Monitor usage**: Keep track of your API usage and costs\n\n---\n\n## 📈 API Key Information\n\n- **Provider**: Moonshot AI (月之暗面)\n- **Endpoint**: https://api.moonshot.cn/anthropic\n- **Documentation**: [Moonshot API Docs](https://platform.moonshot.cn/docs)\n- **Pricing**: Check current rates on their platform\n\n---\n\n## 🎉 Success Indicators\n\n✅ Claude Code installed successfully  \n✅ Moonshot API key configured  \n✅ Organization authentication bypassed  \n✅ Custom endpoint working  \n✅ GitHub repository created  \n✅ Jailbreak environment functional  \n✅ No subscription or organization required  \n\n---\n\n## 📱 Repository Structure\n\n```\nclaude-code-jailbreak/\n├── README.md                     # This tutorial\n├── claude-env.sh                 # Main jailbreak script\n├── .claude-jailbreak-config      # Configuration file\n├── .gitignore                    # Git ignore patterns\n├── LICENSE                       # MIT License\n└── examples/                     # Usage examples (optional)\n```\n\n---\n\n## 🤝 Contributing\n\n1. Fork the repository\n2. Create a feature branch: `git checkout -b feature/amazing-feature`\n3. Commit changes: `git commit -m 'Add amazing feature'`\n4. Push to branch: `git push origin feature/amazing-feature`\n5. Open a Pull Request\n\n---\n\n## ⚖️ Legal Notice\n\nThis tutorial is for educational purposes. Always comply with:\n- Anthropic's Terms of Service\n- Moonshot AI's Terms of Service  \n- Your local laws and regulations\n\n---\n\n## 📞 Support\n\n- **Issues**: [GitHub Issues](https://github.com/chenxingqiang/claude-code-jailbreak/issues)\n- **Discussions**: [GitHub Discussions](https://github.com/chenxingqiang/claude-code-jailbreak/discussions)\n- **Repository**: [GitHub Repository](https://github.com/chenxingqiang/claude-code-jailbreak)\n\n---\n\n**🎯 Goal Achieved**: Claude CLI now works without organization registration using your personal Moonshot API key!\n"}, "5fc7adf5ab47d9fe698c0ef636dbc9c9c6f40577": {"t": 1785765139.4169006, "n": 9845, "h": "84c0ab7a6e15df00", "c": "# 🚀 Claude Code Jailbreak Tutorial\n\n**Complete guide to bypass Claude CLI organization restrictions using Moonshot API**\n\nThis comprehensive tutorial shows you how to install Claude Code, bypass organization authentication, and use your own Moonshot API key instead.\n\n---\n\n## 📋 Table of Contents\n\n1. [Prerequisites](#prerequisites)\n2. [Step 1: Install Claude Code](#step-1-install-claude-code)\n3. [Step 2: Get Moonshot API Key](#step-2-get-moonshot-api-key)\n4. [Step 3: Jailbreak Setup](#step-3-jailbreak-setup)\n5. [Step 4: Create GitHub Repository](#step-4-create-github-repository)\n6. [Step 5: Testing & Verification](#step-5-testing--verification)\n7. [Usage Examples](#usage-examples)\n8. [Troubleshooting](#troubleshooting)\n\n---\n\n## Prerequisites\n\n- **macOS/Linux/Windows** with terminal access\n- **Node.js** (v16 or higher) - [Download here](https://nodejs.org/)\n- **Git** - [Download here](https://git-scm.com/)\n- **GitHub account** - [Sign up here](https://github.com/)\n\n---\n\n## Step 1: Install Claude Code\n\n### Option A: Direct Installation (Recommended)\n```bash\n# Install Claude Code globally\nnpm install -g @anthropic-ai/claude-code\n\n# Verify installation\nclaude --version\n```\n\n### Option B: Alternative Installation Methods\n```bash\n# Using curl (Linux/macOS)\ncurl -fsSL https://claude.ai/claude-code/install.sh | sh\n\n# Using PowerShell (Windows)\nirm https://claude.ai/claude-code/install.ps1 | iex\n```\n\n### Verify Installation\n```bash\n# Check if Claude is installed correctly\nwhich claude\nclaude --help\n```\n\n---\n\n## Step 2: Get Moonshot API Key\n\n### 2.1 Register at Moonshot AI\n1. Visit [Moonshot AI Console](https://platform.moonshot.cn/)\n2. Create an account or sign in\n3. Navigate to API Keys section\n4. Generate a new API key\n\n### 2.2 API Key Format\nYour key should look like: `sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx`\n\n**🔒 Security Note**: Never share your API key publicly!\n\n---\n\n## Step 3: Jailbreak Setup\n\n### 3.1 Create Project Directory\n```bash\n# Create and navigate to project directory\nmkdir claude-code-jailbreak\ncd claude-code-jailbreak\n\n# Initialize git repository\ngit init\n```\n\n### 3.2 Create Environment Configuration\n\nCreate the jailbreak environment file:\n\n```bash\ncat > claude-env.sh << 'EOF'\n#!/bin/bash\n# Claude CLI Jailbreak Configuration for Moonshot API\n# This bypasses organization authentication\n\n# Decode base64 encoded API key (security through obscurity)\nENCODED_KEY=\"c2stNTBmR2I2ZzlYN2w2N1ZOQzFKQ0VoMWtjclFJQmVnbEhyUHN6MDlSSTVmWGcwSVQx\"\nDECODED_KEY=$(echo $ENCODED_KEY | base64 -d)\n\n# Set environment variables for Claude CLI\nexport ANTHROPIC_API_KEY=\"$DECODED_KEY\"\nexport ANTHROPIC_BASE_URL=\"https://api.moonshot.cn/anthropic\"\n\n# Alternative variables for compatibility\nexport MOONSHOT_API_KEY=\"$DECODED_KEY\"\nexport MOONSHOT_BASE_URL=\"https://api.moonshot.cn/anthropic\"\nexport ANTHROPIC_AUTH_TOKEN=\"$DECODED_KEY\"\n\n# Success message\necho \"🔓 Claude CLI Jailbreak Activated!\"\necho \"🔑 API Key: ${ANTHROPIC_API_KEY:0:10}...\"\necho \"🌐 Endpoint: $ANTHROPIC_BASE_URL\"\necho \"✅ Ready to bypass organization restrictions!\"\necho \"\"\necho \"Usage: claude --print 'your prompt here'\"\nEOF\n\n# Make executable\nchmod +x claude-env.sh\n```\n\n### 3.3 Create Configuration File\n```bash\ncat > .claude-jailbreak-config << 'EOF'\n{\n  \"version\": \"1.0.0\",\n  \"provider\": \"moonshot\",\n  \"bypass_auth\": true,\n  \"api_endpoint\": \"https://api.moonshot.cn/anthropic\",\n  \"features\": {\n    \"no_organization_required\": true,\n    \"custom_api_key\": true,\n    \"bypass_restrictions\": true\n  }\n}\nEOF\n```\n\n---\n\n## Step 4: Create GitHub Repository\n\n### 4.1 Create Repository on GitHub\n```bash\n# Login to GitHub CLI (if not already)\ngh auth login\n\n# Create repository\ngh repo create claude-code-jailbreak --public --description \"Claude CLI jailbreak using Moonshot API to bypass organization restrictions\"\n\n# Add remote origin\ngit remote add origin https://github.com/YOUR_USERNAME/claude-code-jailbreak.git\n```\n\n### 4.2 Create Additional Files\n\n**Create .gitignore:**\n```bash\ncat > .gitignore << 'EOF'\n# Security - Never commit real API keys\n*.env\n.env.*\n**/secrets/**\n**/*secret*\n**/*key*\n\n# Logs\n*.log\nlogs/\n\n# OS Files\n.DS_Store\nThumbs.db\n\n# IDE\n.vscode/\n.idea/\n*.swp\n*.swo\n\n# Dependencies\nnode_modules/\nEOF\n```\n\n**Create LICENSE:**\n```bash\ncat > LICENSE << 'EOF'\nMIT License\n\nCopyright (c) 2024 Claude Code Jailbreak\n\nPermission is hereby granted, free of charge, to any person obtaining a copy\nof this software and associated documentation files (the \"Software\"), to deal\nin the Software without restriction, including without limitation the rights\nto use, copy, modify, merge, publish, distribute, sublicense, and/or sell\ncopies of the Software, and to permit persons to whom the Software is\nfurnished to do so, subject to the following conditions:\n\nThe above copyright notice and this permission notice shall be included in all\ncopies or substantial portions of the Software.\n\nTHE SOFTWARE IS PROVIDED \"AS IS\", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR\nIMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,\nFITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE\nAUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER\nLIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,\nOUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE\nSOFTWARE.\nEOF\n```\n\n### 4.3 Commit and Push\n```bash\n# Stage all files\ngit add .\n\n# Commit changes\ngit commit -m \"🚀 Initial commit: Claude Code jailbreak setup with Moonshot API\"\n\n# Push to GitHub\ngit push -u origin main\n```\n\n---\n\n## Step 5: Testing & Verification\n\n### 5.1 Test the Jailbreak\n```bash\n# Activate jailbreak environment\nsource claude-env.sh\n\n# Test basic functionality\nclaude --print \"Hello! Are you working without organization authentication?\"\n\n# Test with more complex prompt\nclaude --print \"Write a simple Python function to calculate fibonacci numbers\"\n```\n\n### 5.2 Verify Success Indicators\nYou should see:\n- ✅ No organization authentication errors\n- ✅ Successful API responses\n- ✅ Custom endpoint being used\n- ✅ API key properly loaded\n\n---\n\n## Usage Examples\n\n### Quick Start\n```bash\n# Every time you want to use Claude CLI:\nsource claude-env.sh\nclaude --print \"Your prompt here\"\n```\n\n### Interactive Mode\n```bash\nsource claude-env.sh\nclaude\n# Now you can chat interactively!\n```\n\n### Advanced Usage\n```bash\n# With specific model\nclaude --print --model sonnet \"Explain quantum computing\"\n\n# With output formatting\nclaude --print --output-format json \"List 5 programming languages\"\n\n# Continue conversation\nclaude --continue\n```\n\n### Permanent Setup (Optional)\nAdd to your `~/.zshrc` or `~/.bashrc`:\n```bash\n# Auto-load Claude jailbreak on terminal start\nif [ -f \"$HOME/claude-code-jailbreak/claude-env.sh\" ]; then\n    source \"$HOME/claude-code-jailbreak/claude-env.sh\"\nfi\n```\n\n---\n\n## 🔧 Troubleshooting\n\n### Common Issues\n\n**Error: \"This organization has been disabled\"**\n- ✅ **Solution**: Use this jailbreak! It bypasses organization auth entirely.\n\n**Error: \"API key not found\"**\n- Check if environment variables are set: `echo $ANTHROPIC_API_KEY`\n- Re-run: `source claude-env.sh`\n- Verify your Moonshot API key is valid\n\n**Error: \"Connection refused\"**\n- Check your internet connection\n- Verify Moonshot API endpoint is accessible\n- Ensure your API key has sufficient credits\n\n**Error: \"Invalid base64 encoding\"**\n- The encoded key might be corrupted\n- Replace with your actual API key in the script\n\n### Debug Mode\n```bash\n# Enable debug mode for detailed error information\nclaude --debug --print \"test prompt\"\n```\n\n### Manual Configuration\nIf the automated setup doesn't work, manually set:\n```bash\nexport ANTHROPIC_API_KEY=\"your-actual-moonshot-api-key\"\nexport ANTHROPIC_BASE_URL=\"https://api.moonshot.cn/anthropic\"\n```\n\n---\n\n## 🔒 Security Notes\n\n1. **API Key Encoding**: The tutorial uses base64 encoding for basic obfuscation\n2. **Never commit real keys**: Always use environment variables or encoded formats\n3. **Regular rotation**: Rotate your API keys periodically\n4. **Monitor usage**: Keep track of your API usage and costs\n\n---\n\n## 📈 API Key Information\n\n- **Provider**: Moonshot AI (月之暗面)\n- **Endpoint**: https://api.moonshot.cn/anthropic\n- **Documentation**: [Moonshot API Docs](https://platform.moonshot.cn/docs)\n- **Pricing**: Check current rates on their platform\n\n---\n\n## 🎉 Success Indicators\n\n✅ Claude Code installed successfully  \n✅ Moonshot API key configured  \n✅ Organization authentication bypassed  \n✅ Custom endpoint working  \n✅ GitHub repository created  \n✅ Jailbreak environment functional  \n✅ No subscription or organization required  \n\n---\n\n## 📱 Repository Structure\n\n```\nclaude-code-jailbreak/\n├── README.md                     # This tutorial\n├── claude-env.sh                 # Main jailbreak script\n├── .claude-jailbreak-config      # Configuration file\n├── .gitignore                    # Git ignore patterns\n├── LICENSE                       # MIT License\n└── examples/                     # Usage examples (optional)\n```\n\n---\n\n## 🤝 Contributing\n\n1. Fork the repository\n2. Create a feature branch: `git checkout -b feature/amazing-feature`\n3. Commit changes: `git commit -m 'Add amazing feature'`\n4. Push to branch: `git push origin feature/amazing-feature`\n5. Open a Pull Request\n\n---\n\n## ⚖️ Legal Notice\n\nThis tutorial is for educational purposes. Always comply with:\n- Anthropic's Terms of Service\n- Moonshot AI's Terms of Service  \n- Your local laws and regulations\n\n---\n\n## 📞 Support\n\n- **Issues**: [GitHub Issues](https://github.com/YOUR_USERNAME/claude-code-jailbreak/issues)\n- **Discussions**: [GitHub Discussions](https://github.com/YOUR_USERNAME/claude-code-jailbreak/discussions)\n- **Email**: Replace with your contact info\n\n---\n\n**🎯 Goal Achieved**: Claude CLI now works without organization registration using your personal Moonshot API key!\n"}}}}