GPAC Test Suite MCP Server
Provides intelligent access to GPAC multimedia framework's 8,600+ test cases, enabling AI assistants to answer questions about GPAC usage with precise, validated command-line examples. Users can ask natural language questions and get reliable GPAC commands for multimedia processing tasks like DASH streaming, MP4 encryption, and subtitle handling.
README
GPAC Test Suite MCP Server
A Model Context Protocol (MCP) server that provides intelligent access to the GPAC multimedia framework test suite, enabling AI assistants to understand and generate GPAC command examples based on validated test cases.
Overview
This MCP server transforms GPAC's extensive test suite (8,600+ test scripts) into an intelligent assistant that can answer questions about GPAC usage with precise, validated examples. Instead of manually searching through thousands of test files, developers can ask natural language questions and get reliable command-line examples.
Key Features
- Semantic Test Search: Find relevant GPAC tests by keywords and descriptions
- Command Generation: Get exact command lines from validated test cases
- Script Analysis: Access and analyze shell test scripts with context
- No Hallucination: Responses based entirely on real, tested GPAC commands
Quick Start with Claude Code
Prerequisites
- Node.js 18+ and pnpm
- Claude Desktop app
- Basic familiarity with MCP configuration
1. Clone and Build
git clone <repository-url>
cd testsuite-mcp
pnpm install
pnpm run build
2. Configure Claude Desktop
Add this MCP server to your Claude Desktop configuration file:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
Linux: ~/.config/Claude/claude_desktop_config.json
{
"mcpServers": {
"gpac-testsuite": {
"command": "node",
"args": ["/absolute/path/to/testsuite-mcp/dist/index.js"],
"env": {
"XML_TESTS_PATH": "/absolute/path/to/testsuite-mcp/all_tests_desriptions.xml",
"SCRIPTS_DIR": "/absolute/path/to/testsuite-mcp/scripts"
}
}
}
}
Important: Use absolute paths in the configuration.
3. Restart Claude Desktop
Close and reopen Claude Desktop for the configuration to take effect.
4. Test the Integration
In Claude Desktop, ask questions like:
- "Show me how to create DASH streaming with GPAC"
- "How to encrypt MP4 files for streaming?"
- "Give me examples of subtitle handling in GPAC"
Development
Project Structure
├── index.ts # Main MCP server entry point
├── xml-tests.ts # XML test descriptions parser
├── scripts-index.ts # Shell scripts indexer
├── all_tests_desriptions.xml # Structured test metadata
├── scripts/ # GPAC test shell scripts
└── dist/ # Compiled JavaScript output
Available MCP Tools
list_xml_tests- List all available XML test namesget_xml_test- Get complete test details with commandsfind_tests_by_keywords- Search tests by keywordsdry_run_xml_test- Generate bash script for test reproductionread_script- Read specific shell script contentfind_commands_by_goal- Find commands by describing the goal
Build Commands
# Development with hot reload
pnpm run dev
# Build for production
pnpm run build
# Run compiled version
pnpm start
Environment Variables
XML_TESTS_PATH: Path to the XML test descriptions file (default:./all_tests_desriptions.xml)SCRIPTS_DIR: Directory containing shell test scripts (default:./scripts)
Usage Examples
Basic Questions
Question: "How to create DASH streaming?"
Response: The assistant will search through test cases and provide exact commands like:
MP4Box -dash 2000 -frag 1000 -profile live input.mp4 -out manifest.mpd
Advanced Workflows
Question: "Can I convert DASH to HLS with encryption without re-segmentation?"
Response: Based on test dash_fwd_hls_twores, you'll get:
gpac -i input.mpd:forward=segb cecrypt:cfile=encryption_config.xml @ -o output.m3u8
Technical Architecture
Data Sources
- XML Test Descriptions (
all_tests_desriptions.xml): Structured metadata describing test purposes, keywords, and expected outcomes - Shell Scripts (
scripts/directory): Actual executable test scripts with environment setup
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。