ieee754-toolbox
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@ieee754-toolbox把 10.2 转成 float 和 double 的十六进制机器码"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
ieee754-toolbox
一个面向逆向工程 / 二进制分析场景的 IEEE 754 浮点工具箱 MCP 服务器。
让支持 MCP 的 AI 助手(WorkBuddy、Claude Desktop、Cursor 等)能够精确完成浮点常量的编码、解码与逐位拆解——patch 二进制浮点常量时不再靠手算。
工具一览
工具 | 功能 | 典型场景 |
| 十进制数 → float / double 十六进制编码 | patch 前计算目标值的机器码(如 10.2 → |
| 十六进制 → 十进制浮点值 | 确认反汇编中立即数对应的浮点值 |
| 逐位拆解符号位 / 指数 / 尾数字段 | 深入理解 IEEE 754 编码结构(指数偏移、隐含前导位) |
Related MCP server: ImHex MCP Integration
环境要求
Python >= 3.10
mcp>= 2.0(注意:mcp 2.x 中FastMCP已改名为MCPServer,本项目使用 2.x API)
安装
pip install -r requirements.txt配置 MCP 客户端
在客户端的 MCP 配置文件中添加(WorkBuddy 为 ~/.workbuddy/mcp.json,Claude Desktop 为 claude_desktop_config.json):
{
"mcpServers": {
"ieee754-toolbox": {
"command": "python",
"args": ["C:/path/to/ieee754-toolbox/server.py"]
}
}
}将 command 换成你的 Python 解释器绝对路径(建议使用虚拟环境),args 换成 server.py 的实际路径。配置后重启客户端,并在连接器管理页面信任该服务器。
冒烟测试(不依赖客户端)
直接用管道向服务器发送 JSON-RPC 消息验证:
printf '%s\n' \
'{"jsonrpc":"2.0","id":0,"method":"initialize","params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"smoketest","version":"1.0"}}}' \
'{"jsonrpc":"2.0","method":"notifications/initialized"}' \
'{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}' \
'{"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"float_to_hex","arguments":{"value":10.2}}}' \
| python server.py预期输出:initialize 响应(含 serverInfo)、3 个工具的清单、float_to_hex 返回 0x41233333 与 0x4024666666666666。
使用示例
配置完成后,直接对 AI 助手说自然语言即可:
"把 3.14 编成 double 的机器码" → 调用
float_to_hex"0x4024666666666666 是多少" → 调用
hex_to_float"拆解一下 0x41233333 这个 float 的字段" → 调用
float_bit_breakdown
背景知识
IEEE 754 浮点数由符号位、指数(带偏移量:float=127,double=1023)、尾数(含隐含前导 1)组成。同一个十进制数在 float(32位)与 double(64位)中编码完全不同——本工具箱帮你免手算处理两种格式。
License
MIT
This server cannot be deployed
Maintenance
Related MCP Connectors
Use AI models for chat, image, and video generation from Claude Code and other MCP hosts.
AI-callable calculators and engineering models with real formulas. No hallucinated math.
Reasoning, code, anti-deception, memory harness MCP tools. Stdio or HTTPS api.ejentum.com/mcp
Your memory, everywhere AI goes. Build knowledge once, access it via MCP anywhere.
Related MCP Servers
- AlicenseNot gradedqualityFmaintenanceA FastMCP server plugin for the Hopper disassembler that enables AI assistants to analyze binary files, disassemble procedures, generate call graphs, search strings, and manage reverse engineering tasks. Provides comprehensive binary analysis capabilities including decompilation, annotation tools, and reference tracking through natural language interactions.81MIT
- FlicenseNot gradedqualityFmaintenanceEnables AI assistants to perform autonomous binary analysis, malware inspection, firmware analysis, and reverse engineering workflows through a production-ready Python MCP server connected to the ImHex hex editor.23-
- FlicenseNot gradedqualityDmaintenanceAn MCP server providing deterministic base conversions (binary, hex, octal, decimal, base64) for AI agents, eliminating hallucination risks in low-level data processing.10 npm3-
- -licenseNot gradedqualityNot gradedmaintenanceExposes Verificarlo floating-point debugging capabilities to AI agents, enabling compilation, instrumentation, and analysis of numerical code through tool calls over the MCP protocol.-