PT5 MCP Server
Enables parsing, analysis, and export of power measurement data from Monsoon Power Monitor PT5 files, including summary statistics, sample retrieval, trend analysis, and CSV export.
README
PT5 MCP Server
MCP (Model Context Protocol) server for parsing Monsoon Power Monitor PT5 files.
Enables any MCP-compatible client (WorkBuddy, Claude Desktop, Cursor, etc.) to read, analyze, and export power measurement data from .pt5 files captured by Monsoon Solutions' Power Tool software.
Features
- parse_pt5 — Parse a PT5 file and return summary statistics (avg/min/max current, voltage, power, energy)
- get_pt5_channels — List available measurement channels and field types from the captureDataMask
- get_pt5_samples — Retrieve sample data with time-range filtering and decimation
- export_pt5_csv — Export sample data to CSV file for external analysis
- get_pt5_header — Return raw header and status packet metadata (hardware info, calibration, scaling)
- analyze_pt5_trend — Analyze current trend: peak detection, periodicity (autocorrelation), trend segments, anomalies
Quick Start
Use with npx (no install needed)
Add to your MCP client configuration (e.g., ~/.workbuddy/mcp.json or Claude Desktop claude_desktop_config.json):
{
"mcpServers": {
"pt5": {
"command": "npx",
"args": ["pt5-mcp-server"]
}
}
}
Install globally
npm install -g pt5-mcp-server
Then configure:
{
"mcpServers": {
"pt5": {
"command": "pt5-mcp-server"
}
}
}
Local development
git clone https://github.com/YOUR_USERNAME/pt5-mcp-server.git
cd pt5-mcp-server
npm install
npm run build
Configure with local path:
{
"mcpServers": {
"pt5": {
"command": "node",
"args": ["/path/to/pt5-mcp-server/build/index.js"]
}
}
}
PT5 File Format
PT5 is the native binary format of Monsoon Solutions' Power Tool software. The file structure is:
| Section | Offset | Description |
|---|---|---|
| Header | 0 | 212-byte fixed header with capture metadata |
| Status Packet | 272 | Variable-length hardware status (calibration, scaling) |
| Sample Data | 1024 | Sequential current/voltage samples at 5 kHz |
Each sample contains:
- Main Current (signed Int32, µA) — if channel enabled
- USB Current (signed Int32, µA) — if channel enabled
- Aux Current (signed Int32, µA) — if channel enabled
- Voltage (unsigned UInt16, with marker bits) — always present
Voltage tick resolution depends on hardware revision:
| Hardware | Main | USB | Aux |
|---|---|---|---|
| Rev A | 62.5 µV/tick | 62.5 µV/tick | 62.5 µV/tick |
| Rev B | 125 µV/tick | 125 µV/tick | 62.5 µV/tick |
| Rev C | 125 µV/tick | 125 µV/tick | 125 µV/tick |
| HVPM | 500 µV/tick | 125 µV/tick | 500 µV/tick |
Tools Reference
parse_pt5
Parse a PT5 file and return summary statistics.
Parameters:
file_path(string, required): Absolute path to the .pt5 file
Returns: JSON with sample count, duration, avg/min/max current (mA), avg voltage (V), avg power (mW), total energy (mJ).
get_pt5_channels
List available measurement channels and fields.
Parameters:
file_path(string, required): Absolute path to the .pt5 file
Returns: Channels (Main/USB/Aux), fields (Min/Avg/Max Voltage/Current/Power), sample rate, hardware info.
get_pt5_samples
Retrieve sample data with filtering options.
Parameters:
file_path(string, required): Absolute path to the .pt5 filestart_time_sec(number, optional): Start time in seconds (default: 0)end_time_sec(number, optional): End time in seconds (default: end of capture)decimation(number, optional): Return every Nth sample (default: 1)max_samples(number, optional): Max samples to return (default: 5000)
Returns: Array of sample objects with timestamp, current (mA), and voltage (V).
export_pt5_csv
Export sample data to a CSV file.
Parameters:
input_path(string, required): Absolute path to the input .pt5 fileoutput_path(string, required): Absolute path for the output .csv file
Returns: Confirmation with file size and sample count.
get_pt5_header
Return raw header and status packet metadata.
Parameters:
file_path(string, required): Absolute path to the .pt5 file
Returns: Full header and status packet data, including voltage tick sizes per channel.
analyze_pt5_trend
Analyze current trend in a PT5 file: detect periodic peaks, autocorrelation-based periodicity, trend direction, and anomalies.
Parameters:
file_path(string, required): Absolute path to the .pt5 filechannel(enum: "main" | "usb" | "aux", optional): Channel to analyze (default: "main")peak_window_ms(number, optional): Peak detection window in ms (default: 50)min_peak_prominence_ma(number, optional): Minimum peak prominence in mA (default: auto — 5% of mean current)anomaly_threshold_std(number, optional): Anomaly detection threshold in standard deviations (default: 3)max_lag_sec(number, optional): Maximum lag for autocorrelation in seconds (default: 10)
Returns:
- Basic stats: mean, std, CV, min, max current
- Peak analysis: total peaks, top peaks with time/current/prominence, average peak interval
- Periodicity: dominant period/frequency from autocorrelation, confidence score (0–1)
- Trend: overall direction (rising/falling/stable), slope, segment breakdown
- Anomalies: intervals where current deviates >N standard deviations from mean
- Histogram: current distribution bins
Example output:
=== Trend Analysis: air-mode.pt5 (channel: main) ===
Mean current: 131.854 mA, Std: 82.512 mA, CV: 0.6259
--- Peak Analysis ---
Total peaks detected: 4244
Average peak interval: 0.0531s (std: 0.0073s)
Top peaks: t=4.610s I=3783.4mA, t=4.709s I=2527.3mA ...
--- Periodicity (Autocorrelation) ---
Dominant period: 0.052s (19.2308Hz)
Confidence: 57.3%
--- Overall Trend ---
Trend: stable (slope: -0.0172 mA/s)
Segments: 80 rising, 60 falling, 84 stable (224 total)
--- Anomalies ---
Found 35 anomaly intervals
License
MIT
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。
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。
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。
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。
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。
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。
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。
mcp-server-qdrant
这个仓库展示了如何为向量搜索引擎 Qdrant 创建一个 MCP (Managed Control Plane) 服务器的示例。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。