Skip to main content
Glama
Sign In
profili-mcp

profili-mcp

by GaillumXc
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Local

profili-mcp

中文 | English

老牌翼型软件 Profili 2.x 的 MCP 服务器。

Profili 是一款经典的翼型设计与分析软件（内置 2200+ 翼型、126 万条极曲线数据），开发已于 2015 年前后停止。它没有脚本 API，也没有 COM/OLE 自动化接口，所以本项目转而与它实际存在的三个接口打交道：

  • Jet 数据库（Profili.mdb）——可读写，GUI 运行时同样可以访问

  • 内嵌的 XFOIL 求解器——与 Profili 自身调用方式完全相同的批处理流程

  • 应用程序进程——启动 / 关闭 / 状态检测

以上全部封装为 14 个 MCP 工具。据我所知这是首个面向 Profili 的 MCP 服务器；使这一切成为可能的存储格式逆向结论记录在 docs/SCHEMA.md 中。

NACA 2412 在 Re=30 万下的极曲线：计算结果与 Profili 库对比

实战示例（examples/）：NACA 2412，Re=3×10⁵，计算极曲线与软件内置库极曲线对比。

工具一览

工具

功能

get_status

探测安装：路径、数据库行数、GUI 状态

launch_profili / close_profili

启动 / 关闭 Profili GUI

list_libraries

列出翼型库

search_airfoils

按名称 / 厚度 / 弯度 / 库筛选翼型

get_airfoil_geometry

元数据 + 上下表面坐标（弦长分数，前缘→后缘）

export_airfoil_dat

导出 .dat 坐标文件

import_airfoil_dat

导入 .dat 进翼型库（自动备份）

list_polars

查询某翼型已有的极曲线雷诺数

get_polar

读取存储极曲线（CL/CD/CM-α，SI 单位）

find_airfoils_by_polar

按指定雷诺数、攻角下的气动性能筛选翼型

analyze_airfoil

用内嵌 XFOIL 现算粘性极曲线，可写回翼型库

analyze_naca

直接计算任意 NACA 4/5 位数翼型，不动数据库

redesign_airfoil

XFOIL GDES TSET 修改最大厚度 / 弯度，可另存新翼型

Related MCP server: Windows-MCP

安装

要求：Windows、已安装的 Profili 2.x（Profili.exe、Profili.mdb、Xfoil.exe 位于同一目录）、以及 32 位 Microsoft Access Driver (*.mdb) ODBC 驱动（Win10/11 通常自带；缺失时安装 MDAC）。

git clone https://github.com/GaillumXc/profili-mcp.git
cd profili-mcp
uv sync          # 或 pip install .

告诉服务器 Profili 的安装位置——环境变量，或作为 install_dir 参数传给任意工具：

set PROFILI_DIR=D:\path\to\ProfiliV2
python -m profili_mcp --selftest     # 全链路自检：数据库 + 数值约定 + XFOIL

MCP 客户端配置

{
  "mcpServers": {
    "profili": {
      "command": "python",
      "args": ["-m", "profili_mcp"],
      "env": { "PROFILI_DIR": "D:\\path\\to\\ProfiliV2" }
    }
  }
}

适用于 Kimi Code、Claude Desktop、Cursor 及任何支持 stdio MCP 的客户端。

安全设计

  • 写库前自动备份 Profili.mdb（同目录时间戳副本）；backup=False 可关闭。

  • GUI 运行时数据库保持可读写，但界面不会自动刷新——导入翼型或保存极曲线后重启 Profili 即可看到。

  • 极曲线写入使用与 Profili 自带库导入完全一致的表结构（PolariHeader + PolariCoordinate）。

示例

examples/ 是一个完整的实战分析（NACA 2412，Re=3×10⁵）：库中取几何 → 计算极曲线 → 提取参数（C_lmax 1.3234、α_stall 15.0°、α₀ −2.19°、C_m0 −0.0499），曲线与内置库存极曲线及教科书数值均吻合。

python examples/analysis_compute.py && python examples/analysis_naca2412.py

相关项目

已知问题

  • 非 ASCII 项目路径 + editable 安装：CPython 以系统 locale 编码读取 .pth 文件，因此当 clone 路径含非 ASCII 字符时 editable 安装会失败。请 clone 到纯英文路径，或以 PYTHONPATH=src python -m profili_mcp 方式运行。

  • 本仓库不包含 Profili 软件本体（商业软件，已停止开发）；项目只与已安装的副本交互。

作者与验证范围

账号所有者（GaillumXc）为无人驾驶航空器系统工程专业本科在读学生，本项目是课程学习过程中的个人作品。全部功能仅在 Profili 2.21 单一版本上验证过；其他版本（2.0/2.2x 各修订、XT、Pro 等）未做测试，行为可能存在差异，欢迎提交 issue 反馈。Kimi K2.8（Moonshot AI 编程助手）参与了逆向分析、代码、文档与示例的编写，提交按 Co-authored-by 惯例署名。

License

MIT，见 LICENSE。Profili 为 devCAD / Stefano Nardone 的商业软件，权利归原作者所有；XFOIL 由 Mark Drela 开发并随 Profili 捆绑分发。本项目仅供个人学习与研究。

Available Tools

14 tools
analyze_airfoilAnalyze AirfoilB

Compute a fresh viscous polar with the bundled XFOIL (same recipe as Profili's own solver). If save_to_db=True the result is written into the Profili library and appears in the GUI after restart.

Inputs: name (required), re, mach, ncrit, +6 more. Returns: result

analyze_nacaAnalyze NacaB

Compute a polar for any NACA 4/5-digit airfoil without touching the library (XFOIL generates the geometry itself).

Inputs: naca_code (required), re, mach, ncrit, +4 more. Returns: result

close_profiliClose ProfiliB

Close all running Profili GUI instances.

Inputs: no named parameters. Returns: result

export_airfoil_datExport Airfoil DatC

Export an airfoil from the library to a Profili .dat coordinate file.

Inputs: name (required), install_dir, output_path. Returns: result

find_airfoils_by_polarFind Airfoils By PolarB

Search airfoils by aerodynamic performance at one operating point.

Narrows by name prefix first (the polar table is huge), then evaluates the polar at alfa (linear interpolation) and applies the criteria.

Inputs: re (required), alfa (required), name_prefix (required), limit, +4 more. Returns: result

get_airfoil_geometryGet Airfoil GeometryB

Get an airfoil's metadata + full coordinates (chord fractions, LE->TE for both surfaces).

Inputs: name (required), install_dir. Returns: result

get_polarGet PolarB

Get the stored polar (CL/CD/CM vs alpha, degrees) at the Reynolds number closest to re.

Inputs: re (required), name (required), install_dir. Returns: result

get_statusGet StatusA

Probe the Profili installation: paths, database row counts, whether the GUI is running, and platform notes. Call this first.

Inputs: install_dir. Returns: result

import_airfoil_datImport Airfoil DatB

Import a .dat coordinate file into the Profili library. The database is backed up first unless backup=False. Restart Profili afterwards if it was already running.

Inputs: file_path (required), name, note, backup, +2 more. Returns: result

launch_profiliLaunch ProfiliC

Start the Profili GUI.

Inputs: install_dir. Returns: result

list_librariesList LibrariesC

List airfoil libraries defined in the Profili database.

Inputs: install_dir. Returns: result

list_polarsList PolarsC

List the Reynolds numbers for which stored polars exist.

Inputs: name (required), install_dir. Returns: result

redesign_airfoilRedesign AirfoilB

Rescale max thickness and/or camber (chord fractions) via XFOIL GDES TSET — the same 'modify airfoil' recipe Profili uses internally. If save_as is given, the new geometry is saved into the library under that name (and as a .dat file next to it).

Inputs: name (required), backup, camber, save_as, +4 more. Returns: result

search_airfoilsSearch AirfoilsC

Search the 2200+ airfoil library. name is a SQL LIKE substring (e.g. 'NACA 24' or 'AG '). Thickness/camber bounds use the database's own units for MaxWidthSuCorda / MaxCamberSuCorda (see docs/SCHEMA.md).

Inputs: name, limit, library, camber_max, +4 more. Returns: result

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Deploy Server
    One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

    TDQS

    B3.4/5.0

    Scored across 14 tools

    Disambiguation4/5

    Most tools target clearly distinct actions: listing, probing, GUI lifecycle, library search, geometry retrieval, import/export, polar lookup, fresh XFOIL analysis, and geometry redesign. The only plausible confusion is between get_polar (stored polar) and analyze_airfoil (fresh computed polar), plus find_airfoils_by_polar vs search_airfoils, but the descriptions make the stored-vs-computed and name-vs-performance distinctions explicit.

    Naming Consistency5/5

    Every tool follows a consistent snake_case verb_noun pattern (list_polars, get_status, search_airfoils, export_airfoil_dat, analyze_naca). Domain nouns are used uniformly, and the launch_/close_ pair is symmetric.

    Tool Count5/5

    14 tools is squarely within the well-scoped range and each tool maps to a distinct capability (inspection, GUI control, library CRUD-ish operations, aerodynamics analysis). No obvious filler or redundant tools.

    Completeness4/5

    The surface covers discovery, inspection, geometry access, import/export, polar analysis, and geometry modification, which covers the main airfoil-database workflows. Notable gaps are deletion/rename of library airfoils and any library-creation operation, though these are safely workable around via import/export.

    Maintenance

    ActivityMaintained
    ResponsivenessNo recent community issues

    Resources

    Related MCP Connectors

    Related MCP Servers

    • Windows-MCP

      StepByStep-1
      A
      license
      Not graded
      quality
      C
      maintenance
      Enables AI agents to interact with Windows operating systems through native UI automation, file navigation, application control, and system commands. Provides seamless integration between LLMs and Windows environments for tasks like clicking, typing, launching apps, and capturing desktop state.
      MIT

    • Windows-MCP

      CyanQX
      A
      license
      Not graded
      quality
      B
      maintenance
      Enables AI agents to automate Windows desktop via human-like mouse and keyboard input and UI Automation, with post-action verification for reliable control of programs, UI elements, windows, and files.
      5
      MIT