local-openai2anthropic 0.2.4__tar.gz → 0.2.7__tar.gz

{local_openai2anthropic-0.2.4 → local_openai2anthropic-0.2.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: local-openai2anthropic
-Version: 0.2.4
+Version: 0.2.7
 Summary: A lightweight proxy server that converts Anthropic Messages API to OpenAI API
 Project-URL: Homepage, https://github.com/dongfangzan/local-openai2anthropic
 Project-URL: Repository, https://github.com/dongfangzan/local-openai2anthropic
@@ -55,6 +55,7 @@ This proxy translates Claude SDK calls to OpenAI API format in real-time, enabli
 - **Offline development** without cloud API costs
 - **Privacy-first AI** - data never leaves your machine
 - **Seamless model switching** between cloud and local
+- **Web Search tool** - built-in Tavily web search for local models
 
 ---
 
@@ -79,7 +80,11 @@ Other OpenAI-compatible backends may work but are not fully tested.
 pip install local-openai2anthropic
 ```
 
-### 2. Start Your Local LLM Server
+### 2. Configure Your LLM Backend (Optional)
+
+**Option A: Start a local LLM server**
+
+If you don't have an LLM server running, you can start one locally:
 
 Example with vLLM:
 ```bash
@@ -93,6 +98,16 @@ sglang launch --model-path meta-llama/Llama-2-7b-chat-hf --port 8000
 # SGLang starts at http://localhost:8000/v1
 ```
 
+**Option B: Use an existing OpenAI-compatible API**
+
+If you already have a deployed OpenAI-compatible API (local or remote), you can use it directly. Just note the base URL for the next step.
+
+Examples:
+- Local vLLM/SGLang: `http://localhost:8000/v1`
+- Remote API: `https://api.example.com/v1`
+
+> **Note:** If you're using [Ollama](https://ollama.com), it natively supports the Anthropic API format, so you don't need this proxy. Just point your Claude SDK directly to `http://localhost:11434/v1`.
+
 ### 3. Start the Proxy
 
 **Option A: Run in background (recommended)**
@@ -155,22 +170,31 @@ You can configure [Claude Code](https://github.com/anthropics/claude-code) to us
 
 ### Configuration Steps
 
-1. **Create or edit Claude Code config file** at `~/.claude/CLAUDE.md`:
-
-```markdown
-# Claude Code Configuration
-
-## API Settings
-
-- Claude API Base URL: http://localhost:8080
-- Claude API Key: dummy-key
-
-## Model Settings
-
-Use model: meta-llama/Llama-2-7b-chat-hf  # Your local model name
+1. **Edit Claude Code config file** at `~/.claude/settings.json`:
+
+```json
+{
+  "env": {
+    "ANTHROPIC_BASE_URL": "http://localhost:8080",
+    "ANTHROPIC_API_KEY": "dummy-key",
+    "ANTHROPIC_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_DEFAULT_SONNET_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_DEFAULT_OPUS_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_REASONING_MODEL": "meta-llama/Llama-2-7b-chat-hf"
+  }
+}
 ```
 
-2. **Alternatively, set environment variables** before running Claude Code:
+| Variable | Description |
+|----------|-------------|
+| `ANTHROPIC_MODEL` | General model setting |
+| `ANTHROPIC_DEFAULT_SONNET_MODEL` | Default model for Sonnet mode (Claude Code default) |
+| `ANTHROPIC_DEFAULT_OPUS_MODEL` | Default model for Opus mode |
+| `ANTHROPIC_DEFAULT_HAIKU_MODEL` | Default model for Haiku mode |
+| `ANTHROPIC_REASONING_MODEL` | Default model for reasoning tasks |
+
+2. **Or set environment variables** before running Claude Code:
 
 ```bash
 export ANTHROPIC_BASE_URL=http://localhost:8080
@@ -179,38 +203,36 @@ export ANTHROPIC_API_KEY=dummy-key
 claude
 ```
 
-3. **Or use the `--api-key` and `--base-url` flags**:
-
-```bash
-claude --api-key dummy-key --base-url http://localhost:8080
-```
-
 ### Complete Workflow Example
 
+Make sure `~/.claude/settings.json` is configured as described above.
+
 Terminal 1 - Start your local LLM:
 ```bash
 vllm serve meta-llama/Llama-2-7b-chat-hf
 ```
 
-Terminal 2 - Start the proxy:
+Terminal 2 - Start the proxy (background mode):
 ```bash
 export OA2A_OPENAI_BASE_URL=http://localhost:8000/v1
 export OA2A_OPENAI_API_KEY=dummy
 export OA2A_TAVILY_API_KEY="tvly-your-tavily-api-key"  # Optional: enable web search
 
-oa2a
+oa2a start
 ```
 
-Terminal 3 - Launch Claude Code with local LLM:
+Terminal 3 - Launch Claude Code:
 ```bash
-export ANTHROPIC_BASE_URL=http://localhost:8080
-export ANTHROPIC_API_KEY=dummy-key
-
 claude
 ```
 
 Now Claude Code will use your local LLM instead of the cloud API.
 
+To stop the proxy:
+```bash
+oa2a stop
+```
+
 ---
 
 ## Features

{local_openai2anthropic-0.2.4 → local_openai2anthropic-0.2.7}/README.md

@@ -20,6 +20,7 @@ This proxy translates Claude SDK calls to OpenAI API format in real-time, enabli
 - **Offline development** without cloud API costs
 - **Privacy-first AI** - data never leaves your machine
 - **Seamless model switching** between cloud and local
+- **Web Search tool** - built-in Tavily web search for local models
 
 ---
 
@@ -44,7 +45,11 @@ Other OpenAI-compatible backends may work but are not fully tested.
 pip install local-openai2anthropic
 ```
 
-### 2. Start Your Local LLM Server
+### 2. Configure Your LLM Backend (Optional)
+
+**Option A: Start a local LLM server**
+
+If you don't have an LLM server running, you can start one locally:
 
 Example with vLLM:
 ```bash
@@ -58,6 +63,16 @@ sglang launch --model-path meta-llama/Llama-2-7b-chat-hf --port 8000
 # SGLang starts at http://localhost:8000/v1
 ```
 
+**Option B: Use an existing OpenAI-compatible API**
+
+If you already have a deployed OpenAI-compatible API (local or remote), you can use it directly. Just note the base URL for the next step.
+
+Examples:
+- Local vLLM/SGLang: `http://localhost:8000/v1`
+- Remote API: `https://api.example.com/v1`
+
+> **Note:** If you're using [Ollama](https://ollama.com), it natively supports the Anthropic API format, so you don't need this proxy. Just point your Claude SDK directly to `http://localhost:11434/v1`.
+
 ### 3. Start the Proxy
 
 **Option A: Run in background (recommended)**
@@ -120,22 +135,31 @@ You can configure [Claude Code](https://github.com/anthropics/claude-code) to us
 
 ### Configuration Steps
 
-1. **Create or edit Claude Code config file** at `~/.claude/CLAUDE.md`:
-
-```markdown
-# Claude Code Configuration
-
-## API Settings
-
-- Claude API Base URL: http://localhost:8080
-- Claude API Key: dummy-key
-
-## Model Settings
-
-Use model: meta-llama/Llama-2-7b-chat-hf  # Your local model name
+1. **Edit Claude Code config file** at `~/.claude/settings.json`:
+
+```json
+{
+  "env": {
+    "ANTHROPIC_BASE_URL": "http://localhost:8080",
+    "ANTHROPIC_API_KEY": "dummy-key",
+    "ANTHROPIC_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_DEFAULT_SONNET_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_DEFAULT_OPUS_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_REASONING_MODEL": "meta-llama/Llama-2-7b-chat-hf"
+  }
+}
 ```
 
-2. **Alternatively, set environment variables** before running Claude Code:
+| Variable | Description |
+|----------|-------------|
+| `ANTHROPIC_MODEL` | General model setting |
+| `ANTHROPIC_DEFAULT_SONNET_MODEL` | Default model for Sonnet mode (Claude Code default) |
+| `ANTHROPIC_DEFAULT_OPUS_MODEL` | Default model for Opus mode |
+| `ANTHROPIC_DEFAULT_HAIKU_MODEL` | Default model for Haiku mode |
+| `ANTHROPIC_REASONING_MODEL` | Default model for reasoning tasks |
+
+2. **Or set environment variables** before running Claude Code:
 
 ```bash
 export ANTHROPIC_BASE_URL=http://localhost:8080
@@ -144,38 +168,36 @@ export ANTHROPIC_API_KEY=dummy-key
 claude
 ```
 
-3. **Or use the `--api-key` and `--base-url` flags**:
-
-```bash
-claude --api-key dummy-key --base-url http://localhost:8080
-```
-
 ### Complete Workflow Example
 
+Make sure `~/.claude/settings.json` is configured as described above.
+
 Terminal 1 - Start your local LLM:
 ```bash
 vllm serve meta-llama/Llama-2-7b-chat-hf
 ```
 
-Terminal 2 - Start the proxy:
+Terminal 2 - Start the proxy (background mode):
 ```bash
 export OA2A_OPENAI_BASE_URL=http://localhost:8000/v1
 export OA2A_OPENAI_API_KEY=dummy
 export OA2A_TAVILY_API_KEY="tvly-your-tavily-api-key"  # Optional: enable web search
 
-oa2a
+oa2a start
 ```
 
-Terminal 3 - Launch Claude Code with local LLM:
+Terminal 3 - Launch Claude Code:
 ```bash
-export ANTHROPIC_BASE_URL=http://localhost:8080
-export ANTHROPIC_API_KEY=dummy-key
-
 claude
 ```
 
 Now Claude Code will use your local LLM instead of the cloud API.
 
+To stop the proxy:
+```bash
+oa2a stop
+```
+
 ---
 
 ## Features

{local_openai2anthropic-0.2.4 → local_openai2anthropic-0.2.7}/README_zh.md

@@ -20,6 +20,7 @@
 - **离线开发** - 无需支付云 API 费用
 - **隐私优先** - 数据不出本机
 - **灵活切换** - 云端和本地模型无缝切换
+- **网络搜索** - 内置 Tavily 网页搜索工具，为本地模型提供联网能力
 
 ---
 
@@ -44,7 +45,11 @@
 pip install local-openai2anthropic
 ```
 
-### 2. 启动本地模型服务
+### 2. 配置你的 LLM 后端（可选）
+
+**选项 A：启动本地模型服务**
+
+如果你还没有运行 LLM 服务，可以在本地启动一个：
 
 使用 vLLM 示例：
 ```bash
@@ -58,6 +63,16 @@ sglang launch --model-path meta-llama/Llama-2-7b-chat-hf --port 8000
 # SGLang 在 http://localhost:8000/v1 启动
 ```
 
+**选项 B：使用已有的 OpenAI 兼容 API**
+
+如果你已经部署了 OpenAI 兼容的 API（本地或远程），可以直接使用。记下 base URL 用于下一步。
+
+示例：
+- 本地 vLLM/SGLang：`http://localhost:8000/v1`
+- 远程 API：`https://api.example.com/v1`
+
+> **注意：** 如果你使用 [Ollama](https://ollama.com)，它原生支持 Anthropic API 格式，无需使用本代理工具。直接将 Claude SDK 指向 `http://localhost:11434/v1` 即可。
+
 ### 3. 启动代理
 
 **方式 A: 后台运行（推荐）**
@@ -120,21 +135,30 @@ print(message.content[0].text)
 
 ### 配置步骤
 
-1. **创建或编辑 Claude Code 配置文件** `~/.claude/CLAUDE.md`：
-
-```markdown
-# Claude Code 配置
-
-## API 设置
-
-- Claude API Base URL: http://localhost:8080
-- Claude API Key: dummy-key
-
-## 模型设置
-
-Use model: meta-llama/Llama-2-7b-chat-hf  # 你的本地模型名称
+1. **编辑 Claude Code 配置文件** `~/.claude/settings.json`：
+
+```json
+{
+  "env": {
+    "ANTHROPIC_BASE_URL": "http://localhost:8080",
+    "ANTHROPIC_API_KEY": "dummy-key",
+    "ANTHROPIC_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_DEFAULT_SONNET_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_DEFAULT_OPUS_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_REASONING_MODEL": "meta-llama/Llama-2-7b-chat-hf"
+  }
+}
 ```
 
+| 变量 | 说明 |
+|------|------|
+| `ANTHROPIC_MODEL` | 通用模型配置 |
+| `ANTHROPIC_DEFAULT_SONNET_MODEL` | Sonnet 模式默认模型（Claude Code 默认使用） |
+| `ANTHROPIC_DEFAULT_OPUS_MODEL` | Opus 模式默认模型 |
+| `ANTHROPIC_DEFAULT_HAIKU_MODEL` | Haiku 模式默认模型 |
+| `ANTHROPIC_REASONING_MODEL` | 推理任务默认模型 |
+
 2. **或者在运行 Claude Code 前设置环境变量**：
 
 ```bash
@@ -144,38 +168,36 @@ export ANTHROPIC_API_KEY=dummy-key
 claude
 ```
 
-3. **也可以使用 `--api-key` 和 `--base-url` 参数**：
-
-```bash
-claude --api-key dummy-key --base-url http://localhost:8080
-```
-
 ### 完整工作流示例
 
+确保 `~/.claude/settings.json` 已按上述步骤配置好。
+
 终端 1 - 启动本地模型：
 ```bash
 vllm serve meta-llama/Llama-2-7b-chat-hf
 ```
 
-终端 2 - 启动代理：
+终端 2 - 启动代理（后台运行）：
 ```bash
 export OA2A_OPENAI_BASE_URL=http://localhost:8000/v1
 export OA2A_OPENAI_API_KEY=dummy
 export OA2A_TAVILY_API_KEY="tvly-your-tavily-api-key"  # 可选：启用网页搜索
 
-oa2a
+oa2a start
 ```
 
-终端 3 - 启动 Claude Code 并使用本地模型：
+终端 3 - 启动 Claude Code：
 ```bash
-export ANTHROPIC_BASE_URL=http://localhost:8080
-export ANTHROPIC_API_KEY=dummy-key
-
 claude
 ```
 
 现在 Claude Code 将使用你的本地大模型，而不是云端 API。
 
+如需停止代理：
+```bash
+oa2a stop
+```
+
 ---
 
 ## 功能特性

{local_openai2anthropic-0.2.4 → local_openai2anthropic-0.2.7}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "local-openai2anthropic"
-version = "0.2.4"
+version = "0.2.7"
 description = "A lightweight proxy server that converts Anthropic Messages API to OpenAI API"
 readme = "README.md"
 license = { text = "Apache-2.0" }

{local_openai2anthropic-0.2.4 → local_openai2anthropic-0.2.7}/src/local_openai2anthropic/__init__.py

@@ -3,7 +3,7 @@
 local-openai2anthropic: A proxy server that converts Anthropic Messages API to OpenAI API.
 """
 
-__version__ = "0.2.4"
+__version__ = "0.2.5"
 
 from local_openai2anthropic.protocol import (
     AnthropicError,

{local_openai2anthropic-0.2.4 → local_openai2anthropic-0.2.7}/src/local_openai2anthropic/config.py

@@ -11,34 +11,35 @@ from pydantic_settings import BaseSettings, SettingsConfigDict
 
 class Settings(BaseSettings):
     """Application settings loaded from environment variables."""
-    
+
     model_config = SettingsConfigDict(
         env_prefix="OA2A_",  # OpenAI-to-Anthropic prefix
         env_file=".env",
         env_file_encoding="utf-8",
         case_sensitive=False,
+        extra="ignore",
     )
-    
+
     # OpenAI API Configuration
     openai_api_key: Optional[str] = None
     openai_base_url: str = "https://api.openai.com/v1"
     openai_org_id: Optional[str] = None
     openai_project_id: Optional[str] = None
-    
+
     # Server Configuration
     host: str = "0.0.0.0"
     port: int = 8080
     request_timeout: float = 300.0  # 5 minutes
-    
+
     # API Key for authenticating requests to this server (optional)
     api_key: Optional[str] = None
-    
+
     # CORS settings
     cors_origins: list[str] = ["*"]
     cors_credentials: bool = True
     cors_methods: list[str] = ["*"]
     cors_headers: list[str] = ["*"]
-    
+
     # Logging
     log_level: str = "DEBUG"
 

{local_openai2anthropic-0.2.4 → local_openai2anthropic-0.2.7}/src/local_openai2anthropic/converter.py

@@ -49,11 +49,11 @@ def convert_anthropic_to_openai(
     system = anthropic_params.get("system")
     stop_sequences = anthropic_params.get("stop_sequences")
     stream = anthropic_params.get("stream", False)
-    temperature = anthropic_params.get("temperature")
+    temperature = anthropic_params.get("temperature", 0.6)
     tool_choice = anthropic_params.get("tool_choice")
     tools = anthropic_params.get("tools")
     top_k = anthropic_params.get("top_k")
-    top_p = anthropic_params.get("top_p")
+    top_p = anthropic_params.get("top_p", 0.95)
     thinking = anthropic_params.get("thinking")
     # metadata is accepted but not forwarded to OpenAI
 

{local_openai2anthropic-0.2.4 → local_openai2anthropic-0.2.7}/src/local_openai2anthropic/main.py

@@ -32,7 +32,7 @@ def create_app(settings: Settings | None = None) -> FastAPI:
     app = FastAPI(
         title="local-openai2anthropic",
         description="A proxy server that converts Anthropic Messages API to OpenAI API",
-        version="0.2.0",
+        version="0.2.5",
         docs_url="/docs",
         redoc_url="/redoc",
     )
@@ -182,7 +182,7 @@ Examples:
     parser.add_argument(
         "--version",
         action="version",
-        version="%(prog)s 0.2.0",
+        version="%(prog)s 0.2.5",
     )
 
     # Create subparsers for commands

{local_openai2anthropic-0.2.4 → local_openai2anthropic-0.2.7}/src/local_openai2anthropic/router.py

@@ -406,7 +406,7 @@ async def _handle_with_server_tools(
         async with httpx.AsyncClient(timeout=settings.request_timeout) as client:
             try:
                 # Log full request for debugging
-                logger.info(f"Request body: {json.dumps(params, indent=2, default=str)[:3000]}")
+                logger.debug(f"Request body: {json.dumps(params, indent=2, default=str)[:3000]}")
                 
                 response = await client.post(url, headers=headers, json=params)
 
@@ -421,7 +421,7 @@ async def _handle_with_server_tools(
                     )
 
                 completion_data = response.json()
-                logger.info(f"OpenAI response: {json.dumps(completion_data, indent=2)[:500]}...")
+                logger.debug(f"OpenAI response: {json.dumps(completion_data, indent=2)[:500]}...")
                 from openai.types.chat import ChatCompletion
                 completion = ChatCompletion.model_validate(completion_data)