translation helps MCP proxy

MCP运维云端

提供方 JEdward7777

运维云端 stdio · 本地进程 会话登录 社区 / 第三方 Claude DesktopClaude
查看接入方式

一个Python stdio MCP服务器,作为MCP客户端与translation-helps-mcp服务器之间的桥梁,处理JSON-RPC 2.0格式化和协议转换。

一句话选型结论

按资料字段整理

translation helps MCP proxy 属于运维云端类 MCP 服务器,stdio · 本地进程,会话登录,可直接接入 Claude Desktop、Claude,有可直接执行的安装命令。接入前先确认权限范围;涉及写入、删除或付款的操作,建议先在测试环境验证。

分类运维云端 传输方式stdio · 本地进程 鉴权会话登录 部署方式本地部署
GitHub 星标 0 仓库热度,无公开仓库时显示未标注
工具数 12 暴露给模型的工具数量
资源 / 提示词 0 / 0 资源与提示词模板数量
维护方 社区 / 第三方 核验等级 仅资料核对

工具与能力

运维云端
数量来自来源目录的标注,具体工具清单以官方文档为准
工具数量12暴露给模型可调用的工具个数
资源数量0可读取的资源条目
提示词数量0内置提示词模板数量

能力标签

支持客户端

适用场景

协议转换Python本地部署JSON-RPC翻译工具MCP兼容Python服务器

安装与配置

stdio · 本地进程 会话登录
客户端在本机拉起子进程,数据不出本机
uvx git+https://github.com/JEdward7777/translation_helps_mcp_proxy.git

这条命令只是入口,具体启动参数请按官方文档补齐,不要直接照搬到别的服务器。

  1. 1
    准备运行环境

    stdio 服务在本地起进程,需要先装好 Python,再按安装命令拉取依赖。

  2. 2
    安装服务

    复制上面的安装命令执行;命令里带占位符的位置,换成自己的密钥或路径。

  3. 3
    写入客户端配置

    在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,让新配置生效。

  4. 4
    验证是否可用

    连接成功后客户端会列出约 12 个工具,随便调一个只读工具确认能返回结果。

典型使用场景

3 类
按分类与标签整理的常见用法,具体以官方文档为准
云资源查询

查实例、存储、账单等云上资源状态。

部署与发布

把部署、重启、回滚这类操作接进对话。

监控与告警

拉取监控指标,异常时让模型帮你判断。

选择建议

选之前先看这几条
按条目字段给出的判断,不替你决定用不用

值得选的地方

  • 资料标注 12 个工具,覆盖面较广,能满足多种调用。
  • 有现成的安装命令,不用自己找入口,照着跑就能起服务。
  • 已适配 Claude Desktop、Claude,换客户端不用重新研究配置。

先想清楚的地方

  • stdio 方式要求本地装好运行时(如 Node.js),换机器要重新装一遍。
  • 使用前要准备好会话登录对应的凭据,并按最小权限配置。
  • 版本更新与行为细节以官方仓库为准。

更适合

  • 需要在运维云端场景里让模型直接调用外部能力的人。
  • 已经在用 Claude Desktop、Claude 的用户,接入路径最短。
  • 想先快速试一下、能接受在本地或自有环境跑服务的人。

不太适合

  • 完全不想在本地装环境的人。
  • 不方便管理密钥或第三方授权的人。
  • 需要厂商级 SLA 和技术支持的生产环境。

同类对比

6 个同类
同分类里挑最接近的几条,字段以各自详情页为准
MCP 服务器工具数传输方式部署方式安装命令最后核验
translation helps MCP proxy当前 12 stdio · 本地进程 本地部署 有 2026-05-17
lobehub 未标注 stdio · 本地进程 本地部署 有 2026-04-11
TrendRadar 未标注 stdio · 本地进程 本地部署 有 2026-04-13
Ui Tars Desktop 8a825770 未标注 stdio · 本地进程 本地部署 有 2026-05-17
ruflo 未标注 stdio · 本地进程 本地部署 有 2026-04-13
FastGPT 未标注 stdio · 本地进程 本地部署 有 2026-04-11

服务器档案

运维云端
字段来自项目仓库与公开索引,中文说明由本站标注
提供方JEdward7777上游标注的维护方 JEdward7777
分类运维云端本站按上游分类展示
连接方式stdio · 本地进程客户端在本机拉起子进程,数据不出本机
鉴权方式会话登录远程服务建议用最小权限凭据
维护方社区 / 第三方核验等级 仅资料核对
包名 / 运行时未标注 · Python本地安装时用到的包标识与运行时
部署类型本地部署上游给出的部署形态
最近核验2026-05-17上游最后核验时间

详细介绍

24.6 KB
来自项目仓库与上游目录的原文,未做删改

📚 Upstream Project

Upstream Details:

🎯 Current Status

✅ PRODUCTION READY - All functionality working as of October 2025

  • MCP Initialization: ✅ Proper JSON-RPC 2.0 handshake
  • Tool Discovery: ✅ Discovers all 12 tools from upstream server
  • Tool Execution: ✅ Successfully executes all tools including fetch_scripture
  • Stdio Protocol: ✅ Complete MCP stdio workflow operational
  • Error Handling: ✅ Graceful error handling and cleanup

Test Results: 37/37 comprehensive tests passing (100% operational)

🔧 Technical Architecture

Protocol Flow```

MCP Client ←→ Proxy Server ←→ Upstream Server
(stdio) (stdio) (HTTP)
JSON-RPC 2.0 JSON-RPC 2.0 API endpoints


### Key Features
- **Response Formatting**: Handles various upstream response formats (MCP content, scripture data, wrapped results)
- **Smart Routing**: Routes `fetch_scripture` calls to dedicated `/api/fetch-scripture` endpoint
- **Error Recovery**: Robust async error handling and resource cleanup

## 📁 Project Structure```
translation_helps_mcp_proxy/
├── src/
│   └── translation_helps_mcp_proxy/
│       ├── __init__.py           # Package initialization
│       ├── __main__.py           # Main entry point
│       └── mcp_proxy_server.py   # Main proxy server
├── tests/                        # Comprehensive pytest test suite (37 tests)
│   ├── __init__.py              # Package initialization
│   ├── conftest.py              # Shared test fixtures
│   ├── test_upstream_connectivity.py  # Upstream connection tests
│   ├── test_mcp_protocol.py           # MCP protocol tests
│   ├── test_tool_execution.py         # Tool execution tests
│   ├── test_stdio_workflow.py         # End-to-end workflow tests
│   ├── test_fetch_translation_notes.py # Translation notes tool tests
│   ├── test_get_translation_word.py    # Translation word tool tests
│   ├── test_tool_filtering.py          # Tool filtering tests
│   └── test_strict_translation_notes.py # Strict translation notes tests
├── pyproject.toml            # Python package configuration (uvx compatible)
├── pytest.ini               # Pytest configuration
├── requirements.txt          # Python dependencies (includes pytest)
├── setup.sh / setup.bat      # Cross-platform setup scripts
├── README.md                 # This file
├── CORRECT_mcp_config.json   # Example MCP client configuration
└── .gitignore               # Git ignore rules

⚡ Quick Start

Using uvx (Recommended - Zero Installation)```bash

Install and run with uvx (no local setup needed)

uvx git+https://github.com/JEdward7777/translation_helps_mcp_proxy.git

Or install globally for repeated use

uvx install git+https://github.com/JEdward7777/translation_helps_mcp_proxy.git
translation-helps-mcp-proxy

With custom options

uvx git+https://github.com/JEdward7777/translation_helps_mcp_proxy.git \
--debug \
--enabled-tools "fetch_scripture,fetch_translation_notes,get_system_prompt"

Hide problematic parameters from downstream tools

uvx git+https://github.com/JEdward7777/translation_helps_mcp_proxy.git \
--hide-params "language,organization"

List available tools

uvx git+https://github.com/JEdward7777/translation_helps_mcp_proxy.git --list-tools


**uvx Benefits:**
- ⚡ **Zero setup** - No Python environment management needed
- 🔄 **Always latest** - Automatically pulls latest version from git
- 🧹 **Clean** - No local dependencies or virtual environments
- 🚀 **Fast** - uvx handles all dependency management automatically

### Using uv sync (Alternative to pip)

If you prefer using `uv` for dependency management instead of pip:```bash
# Clone the repository
git clone https://github.com/JEdward7777/translation_helps_mcp_proxy.git
cd translation_helps_mcp_proxy

# Sync dependencies using uv (creates .venv automatically)
uv sync

# Activate the virtual environment
. .venv/bin/activate  # Linux/macOS
# or
.venv\Scripts\activate  # Windows

# Test the package
python -m src.translation_helps_mcp_proxy --help

# Run tests
pytest

uv sync Benefits:
- 🚀 Faster - Much faster dependency resolution and installation
- 🔒 Locked - Uses uv.lock file for reproducible builds
- 📦 Unified - Single tool for all dependency management
- 🎯 Dev dependencies - Automatically installs dev dependencies from [dependency-groups]

Note: The uv sync command reads from pyproject.toml and creates/updates the virtual environment with all required dependencies, including development dependencies like pytest.

Local Development Setup```bash

Clone the repository

git clone https://github.com/JEdward7777/translation_helps_mcp_proxy.git
cd translation_helps_mcp_proxy

Linux/macOS

./setup.sh

Windows

setup.bat

Or manual setup

python3 -m venv venv
. venv/bin/activate # Linux/macOS (note: use dot, not 'source')
pip install -r requirements.txt

Test the package

python -m src.translation_helps_mcp_proxy --help

Run tests

pytest


**Important**: Always activate the virtual environment before running any commands when developing locally. Use `. venv/bin/activate` (with dot) if `source` command is not available.

## 🧪 Testing

### Modern pytest Test Suite (Recommended)
Run all tests:```bash
# Activate venv first (REQUIRED)
. venv/bin/activate  # Use dot if 'source' doesn't work

# Run all tests
pytest

# Run with verbose output
pytest -v

# Run specific test files
pytest tests/test_upstream_connectivity.py
pytest tests/test_tool_execution.py
pytest tests/test_fetch_translation_notes.py
pytest tests/test_get_translation_word.py
pytest tests/test_tool_filtering.py

# Run specific test functions
pytest tests/test_stdio_workflow.py::test_stdio_workflow
pytest tests/test_fetch_translation_notes.py::test_fetch_translation_notes_basic
pytest tests/test_get_translation_word.py::test_get_translation_word_basic
pytest tests/test_tool_filtering.py::test_tool_filtering_specific_tools

# Run tests in parallel (faster)
pytest -n auto

# Run comprehensive test suite (all tests)
pytest tests/ -v

# Run all tool-specific tests
pytest tests/test_fetch_translation_notes.py tests/test_get_translation_word.py tests/test_tool_filtering.py tests/test_parameter_hiding.py -v

🧪 Comprehensive Test Suite

Total Test Coverage: 43 test cases across 7 test files```bash

Run all tests with detailed output

. .venv/bin/activate
pytest tests/ -v

Expected output summary:

tests/test_upstream_connectivity.py: 2 tests ✅

tests/test_mcp_protocol.py: 2 tests ✅

tests/test_tool_execution.py: 4 tests ✅

tests/test_stdio_workflow.py: 2 tests ✅

tests/test_fetch_translation_notes.py: 7 tests ✅

tests/test_get_translation_word.py: 9 tests ✅

tests/test_tool_filtering.py: 9 tests ✅

tests/test_parameter_hiding.py: 6 tests ✅

tests/test_strict_translation_notes.py: 2 tests ✅

= 43 tests total


**Test Categories:**
- **Core Infrastructure** (8 tests) - Protocol, connectivity, basic tool execution
- **Tool Filtering** (9 tests) - Safety controls and rollout management
- **Parameter Hiding** (6 tests) - Parameter filtering for downstream compatibility
- **Upstream Connectivity** (2 tests) - Connection and tool discovery validation

### Simple Test Execution```bash
# Just run pytest - discovers and runs all 37 tests automatically
. venv/bin/activate
pytest

Test Coverage:
1. Upstream Connectivity ✅ - Verifies connection and tool discovery
2. MCP Protocol ✅ - Tests JSON-RPC 2.0 initialization
3. Tool Execution ✅ - Validates tool calls including fetch_scripture
4. Stdio MCP Workflow ✅ - Complete end-to-end John 3:16 test

pytest Expected Output:```
========================= test session starts =========================
collected 16 items

tests/test_upstream_connectivity.py::test_upstream_connectivity PASSED
tests/test_upstream_connectivity.py::test_tools_list_format PASSED
tests/test_mcp_protocol.py::test_mcp_protocol_initialization PASSED
tests/test_mcp_protocol.py::test_mcp_response_format PASSED
tests/test_tool_execution.py::test_tool_execution PASSED
tests/test_tool_execution.py::test_get_system_prompt_tool PASSED
tests/test_tool_execution.py::test_fetch_scripture_tool PASSED
tests/test_stdio_workflow.py::test_stdio_workflow PASSED
tests/test_fetch_translation_notes.py::test_fetch_translation_notes_basic PASSED
tests/test_fetch_translation_notes.py::test_fetch_translation_notes_missing_reference PASSED
tests/test_fetch_translation_notes.py::test_fetch_translation_notes_different_book PASSED
tests/test_fetch_translation_notes.py::test_fetch_translation_notes_default_params PASSED
tests/test_fetch_translation_notes.py::test_fetch_translation_notes_invalid_reference PASSED
tests/test_fetch_translation_notes.py::test_fetch_translation_notes_response_structure PASSED
tests/test_fetch_translation_notes.py::test_fetch_translation_notes_tool_integration PASSED
tests/test_get_translation_word.py::test_get_translation_word_basic PASSED

========================= 16 passed in X.XXs =========================


### pytest Benefits
- **Individual Test Execution**: Run specific tests for focused debugging
- **Parallel Execution**: Faster test runs with `-n auto`
- **Better Output**: Cleaner, more detailed test reports
- **Context Efficiency**: Individual test files are 40-160 lines vs 418 lines

## 🚀 Usage

### MCP Proxy Mode```bash
# Using uvx (recommended)
uvx git+https://github.com/JEdward7777/translation_helps_mcp_proxy.git

# Local development
. venv/bin/activate
python -m src.translation_helps_mcp_proxy

# Local development (after setup)
. venv/bin/activate
translation-helps-mcp-proxy

# With options
uvx git+https://github.com/JEdward7777/translation_helps_mcp_proxy.git \
--debug \
--enabled-tools "fetch_scripture,fetch_translation_notes,get_system_prompt"

Advanced Options```bash

Different upstream URL

translation-helps-mcp-proxy --upstream-url "http://localhost:5173/api/mcp"

Debug mode (detailed logging)

translation-helps-mcp-proxy --debug

Tool filtering (controlled rollout)

translation-helps-mcp-proxy --enabled-tools "fetch_scripture,fetch_translation_notes"

Parameter hiding (hide specific parameters from tool schemas)

translation-helps-mcp-proxy --hide-params "language,organization"

Combined options

translation-helps-mcp-proxy \
--upstream-url "http://localhost:5173/api/mcp" \
--debug \
--enabled-tools "fetch_scripture,fetch_translation_notes" \
--hide-params "language,organization"


### Production Deployment

**uvx Installation (Recommended)**:```bash
# Install globally on server
uvx install git+https://github.com/JEdward7777/translation_helps_mcp_proxy.git

# Run in production
translation-helps-mcp-proxy --enabled-tools "fetch_scripture,fetch_translation_notes,get_system_prompt"

Docker Deployment:```dockerfile
FROM python:3.11-slim
RUN pip install uv
RUN uvx install git+https://github.com/JEdward7777/translation_helps_mcp_proxy.git
CMD ["translation-helps-mcp-proxy"]


### 🛡️ Tool Filtering for Controlled Rollout

Control which MCP tools are available for enhanced safety during testing:```bash
# FIRST: Discover all available tools
python -m src.translation_helps_mcp_proxy --list-tools

# Enable all tools (default behavior)
python -m src.translation_helps_mcp_proxy

# Enable only specific tools (comma-separated, no spaces around commas)
python -m src.translation_helps_mcp_proxy --enabled-tools "fetch_scripture,fetch_translation_notes"

# Enable only fully tested tools
python -m src.translation_helps_mcp_proxy --enabled-tools "fetch_scripture,fetch_translation_notes,get_system_prompt"

# Disable all tools (useful for testing)
python -m src.translation_helps_mcp_proxy --enabled-tools ""

🔒 Parameter Hiding for Downstream Compatibility

Hide specific parameters from tool schemas to solve downstream compatibility issues:```bash

Hide language and organization parameters from all tools

python -m src.translation_helps_mcp_proxy --hide-params "language,organization"

Hide only specific parameters

python -m src.translation_helps_mcp_proxy --hide-params "language"

Combine with tool filtering

python -m src.translation_helps_mcp_proxy \
--enabled-tools "fetch_scripture,fetch_translation_notes" \
--hide-params "language,organization"


**Use Case**: When downstream tools provide invalid values for certain parameters (e.g., incorrect language codes), you can hide those parameters from the tool schema. This prevents the downstream tool from sending those parameters, allowing the upstream server to use its default values instead.

**How it works**:
- Removes specified parameters from the `properties` section of tool input schemas
- Removes hidden parameters from the `required` list if present
- Applies to ALL tools that have the specified parameters
- Case-sensitive: must match parameter names exactly (e.g., `language` not `Language`)```bash
python -m src.translation_helps_mcp_proxy --hide-params "language,organization"

This hides both parameters, so the downstream tool won't send them, and the upstream server will use default values.```bash

Enable filtering of book and chapter notes

python -m src.translation_helps_mcp_proxy --filter-book-chapter-notes

Combine with other options

python -m src.translation_helps_mcp_proxy \
--filter-book-chapter-notes \
--enabled-tools "fetch_translation_notes"


**Example Problem**:
- Fetching notes for John 3:16 returns:
- ✅ Verse-specific notes for John 3:16 (what you need)
- ❌ "Introduction to John" (repeated for every verse in John)
- ❌ "John 3 Chapter Introduction" (repeated for every verse in chapter 3)

**How it works**:
- Automatically filters out notes with `Reference: "front:intro"` (book-level)
- Automatically filters out notes with `Reference: ":intro"` (chapter-level)
- Keeps only verse-specific notes (e.g., `Reference: "3:16"`)
- Updates the response metadata to reflect the filtered count

**Benefits**:
- **Reduces response size**: Eliminates duplicate book/chapter notes
- **Improves clarity**: Only verse-specific notes are returned
- **Maintains compatibility**: Preserves all metadata and response structure

**Before filtering** (John 3:16):```json
{
"items": [
{"Reference": "front:intro", "Note": "# Introduction to John..."},
{"Reference": "3:intro", "Note": "# John 3 Chapter Introduction..."},
{"Reference": "3:16", "Note": "Verse-specific note 1..."},
{"Reference": "3:16", "Note": "Verse-specific note 2..."}
],
"metadata": {"totalCount": 9}
}

After filtering (with --filter-book-chapter-notes):```json
{
"items": [
{"Reference": "3:16", "Note": "Verse-specific note 1..."},
{"Reference": "3:16", "Note": "Verse-specific note 2..."}
],
"metadata": {"totalCount": 7}
}


**Recommended for**:
- Cost-sensitive applications processing many verses
- Any scenario where book/chapter context is not needed per verse

### 📋 Tool Discovery

**List all available tools** before configuring:```bash
# Discover what tools are available from upstream
. venv/bin/activate
python -m src.translation_helps_mcp_proxy --list-tools

Example output:```
📋 Discovering available tools from upstream server...

✅ Found 12 available tools:

  1. get_system_prompt - Get the complete system prompt and constraints
  2. fetch_scripture - Fetch Bible scripture text for a specific reference
  3. fetch_translation_notes - Fetch translation notes for a specific Bible reference
  4. get_languages - Get available languages for translation resources
  5. fetch_translation_questions - Fetch translation questions for a specific Bible reference
  6. browse_translation_words - Browse and search translation words by category or term
  7. get_context - Get contextual information for a Bible reference
  8. extract_references - Extract and parse Bible references from text
  9. fetch_resources - Fetch multiple types of translation resources for a reference
  10. get_words_for_reference - Get translation words specifically linked to a Bible reference
  11. get_translation_word - Get translation words linked to a specific Bible reference
  12. search_resources - Search across multiple resource types for content

💡 Usage: --enabled-tools "tool1,tool2,tool3"
Example: --enabled-tools "fetch_scripture,fetch_translation_notes"


## 🔧 Integration with MCP Clients

The proxy is designed to work seamlessly with MCP clients like KiloCode, Claude Desktop, and other MCP-compatible tools. It provides a standard MCP stdio interface that any MCP client can connect to.

### Example Integration Workflow```
MCP Client → Translation Helps MCP Proxy → Translation Helps Server
↓              ↓                         ↓
JSON-RPC 2.0 → Protocol Translation → Bible Resources

How It Works:
1. MCP Client: Makes standard MCP tool calls via stdio
2. MCP Proxy: Translates requests to HTTP API calls
4. MCP Response: Formats data for seamless client integration

💡 Recommended Workflow:
1. Run --list-tools to see all available tools
3. Test each new tool individually before adding to your enabled list
4. Gradually expand your enabled tools as you verify each one works

Key Features:
- Safe rollout: Only enable tools you've verified work correctly
- Invalid tool detection: Server will error if you specify non-existent tools
- Case-sensitive: Tool names must match exactly (e.g., fetch_scripture not FETCH_SCRIPTURE)
- No whitespace: Use "tool1,tool2" not "tool1, tool2"

Available Tools for Filtering:```
fetch_scripture ✅ Verified working
fetch_translation_notes ✅ Verified working
get_system_prompt ✅ Verified working
fetch_translation_questions 🔄 Needs verification
browse_translation_words 🔄 Needs verification
get_translation_word 🔄 Needs verification
get_words_for_reference 🔄 Needs verification
get_context 🔄 Needs verification
extract_references 🔄 Needs verification
fetch_resources 🔄 Needs verification
search_resources 🔄 Needs verification
get_languages 🔄 Needs verification


**Example Safe Configuration (Only Verified Tools):**```bash
python -m src.translation_helps_mcp_proxy --enabled-tools "fetch_scripture,fetch_translation_notes,get_system_prompt"

MCP Client Configuration

KiloCode Configuration (Recommended)

Step 1: Find your absolute paths and set up the virtual environment:```bash

Navigate to your translation_helps_mcp_proxy directory (replace with your actual path)

cd /absolute/path/to/your/translation_helps_mcp_proxy

Activate virtual environment

. venv/bin/activate

Note these paths for the configuration

pwd # Current directory (use for 'cwd')
which python # Python executable path (use for 'command')


**Step 2**: Add this configuration to your KiloCode MCP settings JSON:```json
{
"translation-helps-mcp-proxy": {
"command": "/absolute/path/to/your/translation_helps_mcp_proxy/venv/bin/python",
"args": ["-m", "src.translation_helps_mcp_proxy"],
"cwd": "/absolute/path/to/your/translation_helps_mcp_proxy",
"env": {
"PYTHONPATH": "/absolute/path/to/your/translation_helps_mcp_proxy"
}
}
}

Replace the paths above with your actual paths:
- The command should point to your venv's Python: YOUR_PATH/venv/bin/python
- All three path fields should use the same base directory

Key Configuration Notes:
- command: Use the venv Python executable (not system python)
- env.PYTHONPATH: Ensures Python can find the modules

Alternative: Generic MCP Client Configuration

For other MCP clients that don't support virtual environments well:```json
{
"translation-helps-mcp-proxy": {
"command": "python",
"args": ["-m", "src.translation_helps_mcp_proxy"],
"cwd": "/absolute/path/to/translation_helps_mcp_proxy"
}
}


#### Verification Steps

**Test the configuration**:
1. Save the JSON configuration in KiloCode
2. Restart KiloCode or reload MCP servers
3. In KiloCode, verify the server appears in the MCP tools list
4. Try calling a tool like `fetch_scripture` with `{"reference": "John 3:16"}`

**Expected behavior**:
- Server should connect without errors
- 12 tools should be available
- `fetch_scripture` should return John 3:16 text

#### Troubleshooting KiloCode Integration

**If the server doesn't start**:```bash
# Test manually first (replace with your actual path)
cd /absolute/path/to/your/translation_helps_mcp_proxy
. venv/bin/activate
python -m src.translation_helps_mcp_proxy --debug

Common issues:
- Wrong Python path: Ensure you're using venv/bin/python, not system python
- Missing dependencies: Run pip install -r requirements.txt in the venv
- Path issues: Use absolute paths for all directory references
- Permissions: Ensure the venv directory is readable

Debug configuration in KiloCode:
- Check KiloCode's MCP server logs for connection errors
- Ensure no other MCP servers are using the same name

🛠️ Available Tools (12 Total)

The proxy automatically discovers and forwards all tools from the upstream server:

Tool Description Status
get_system_prompt Get system prompt and constraints ✅ Working
fetch_scripture Fetch Bible scripture text ✅ Working
get_context Get contextual information ✅ Working
extract_references Parse Bible references from text ✅ Working
fetch_resources Get multiple resource types ✅ Working
get_words_for_reference Get words for specific references ✅ Working
search_resources Search across resource types ✅ Working
get_languages Get available languages ✅ Working

Example: Fetching John 3:16```json

{
"name": "fetch_scripture",
"arguments": {"reference": "John 3:16"}
}


## 🔍 Response Format Handling

The proxy intelligently handles various upstream response formats:

2. **MCP Format**: `{"content": [...]}` → Proper MCP TextContent objects
3. **Wrapped Results**: `{"result": ...}` → Extracts and formats result
4. **Raw JSON**: Any other format → JSON stringification fallback

## 🐛 Troubleshooting

### Import Errors```bash
# Ensure virtual environment is activated (MOST COMMON ISSUE)
. venv/bin/activate  # Use dot if 'source' command doesn't work

# Reinstall dependencies
pip install -r requirements.txt

⚠️ Common Issue: Most problems are caused by forgetting to activate the virtual environment. Always run . venv/bin/activate before any Python commands.

Connection Issues```bash

IMPORTANT: Activate venv first

. venv/bin/activate

Test upstream connectivity

python -c "
import asyncio, httpx
async def test():
client = httpx.AsyncClient(verify=False)
r = await client.get('https://translation-helps-mcp.pages.dev/api/mcp?method=ping')
print(f'Status: {r.status_code}, Response: {r.text[:100]}')
await client.aclose()
asyncio.run(test())
"


### Debug Mode```bash
# IMPORTANT: Activate venv first
. venv/bin/activate

python -m src.translation_helps_mcp_proxy --debug

Shows detailed upstream communication, response parsing, and error details.

Virtual Environment Notes

  • Always activate venv: All Python commands require the virtual environment to be activated
  • Use dot syntax: If source venv/bin/activate doesn't work, use . venv/bin/activate
  • Per-session: You need to activate the venv in each new terminal session
  • Verification: Your prompt should show (venv) when the virtual environment is active

📊 Performance & Reliability

  • Timeout: 30-second timeout for upstream requests
  • Connection: Persistent HTTP client with connection pooling
  • Memory: Minimal memory footprint, suitable for long-running processes
  • SSL: Handles certificate issues by disabling verification for upstream server
  • Async: Proper async/await with TaskGroup error handling

🔮 Technical Details

Key Dependencies```python

mcp>=1.0.0 # Official MCP SDK
httpx>=0.25.0 # Async HTTP client
typing-extensions>=4.0.0 # Type hints support
```

Core Architecture

  • MCPProxyServer: Main proxy class handling MCP protocol and upstream communication
  • Smart Tool Routing: Special handling for fetch_scripture and other tools
  • Response Transformation: Converts various upstream formats to proper MCP responses
  • Error Handling: Graceful async error handling with proper cleanup

Future Enhancements

Potential improvements for advanced use cases:
- [x] Convert to pytest-based test suite for better CI/CD integration
- [ ] Add response caching for frequently called tools
- [ ] Support for additional MCP features (resources, prompts)
- [ ] Metrics and monitoring capabilities
- [ ] Rate limiting and request queuing

📝 License


常见问题

8 问
答案由本站依据条目字段整理,不引用上游原文
translation helps MCP proxy 怎么安装?

在客户端按下面的安装命令接入:uvx git+https://github.com/JEdward7777/translation_helps_mcp_proxy.git。装完在客户端配置里引用同一个包或端点。

translation helps MCP proxy 怎么在 Claude Desktop、Claude 里配置?

它的传输方式是stdio · 本地进程,在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,配置才会生效。

translation helps MCP proxy 收费吗?

本站不代收费用,也没有标注具体价格。是否收费取决于服务本身和背后的接口额度,正式使用前按官方说明确认。

translation helps MCP proxy 需要 API Key 吗?

资料标注需要会话登录认证,使用前准备好对应凭据,并按最小权限配置。

translation helps MCP proxy 支持哪些客户端?

资料标注支持 Claude Desktop、Claude;其他客户端只要能配置 MCP Server,也能用同样的命令或地址接入。

translation helps MCP proxy 能提供哪些能力?

资料显示约 12 个工具;能力标签包括 协议转换、Python、Claude、Python服务器。具体工具清单以官方文档为准。

translation helps MCP proxy 需要什么运行环境?

运行时为 Python;属于本地进程,需要在本机装好运行环境。

translation helps MCP proxy 是本地还是远程?

它是本地进程(stdio),由客户端在本机拉起。传输方式为stdio · 本地进程。

热门专题

按标签挑服务器
全部专题 →

服务器信息以作者仓库与官方文档为准;分类、传输方式与鉴权方式的中文说明为中文整理,最近核验 2026-05-17。