abydym/AstraLogic

Última atividade de origem registrada: 11 de junho de 2026 às 08:22

Catálogo do SkillsMP atualizado: 12 de novembro de 2026 às 12:52

skills coletadas: 1

Estrelas no GitHub: 4
Forks no GitHub: 0

Instalar com um assistente de IA

Copie este prompt para o assistente de IA que você usa.

Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/abydym/AstraLogic.

README do repositório

Arquivo de origem do README: README.md

Ler o README original

Expandir README completo

<div align="center"> # 🛰️ AstraLogic **Agent-Driven Satellite Communication Simulation Framework** *An intelligent bridge between high-fidelity physics simulation and cognitive decision-making agents* [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE) [![Python 3.8+](https://img.shields.io/badge/python-3.8+-blue.svg)](https://www.python.org/) [![AFSIM](https://img.shields.io/badge/engine-AFSIM%20Warlock-orange.svg)]() [English](#english) | [中文](#chinese) </div> --- <a name="english"></a> ## Overview **AstraLogic** is a one-stop satellite communication simulation & decision platform that seamlessly integrates AFSIM (Advanced Framework for Simulation, Integration, and Modeling) execution with intelligent agent-driven analysis within a single web interface. Users interact with the system entirely through a web dashboard — submitting simulation requests via an Agent chat interface, which delegates decision-making to **OpenClaw** (an intelligent agent framework). OpenClaw reasons about orbital mechanics, link budgets, and communication strategies, then orchestrates AFSIM simulations via **Skills** and **MCP servers**. Results are automatically parsed and visualized in the dashboard without ever needing to leave the browser. The architecture consists of: 1. **Frontend** — Streamlit dashboard with Agent chat interface for end-to-end simulation control 2. **Backend** — FastAPI server coordinating agent decisions and simulation state 3. **Decision Engine** — OpenClaw agent with custom Skills for link analysis and frequency optimization 4. **Simulation Bridge** — MCP server protocol for headless AFSIM execution and result ingestion 5. **Data Layer** — AER parser converting AFSIM binary outputs to structured DataFrames ``` ┌────────────────────────────────────────────────────────────────┐ │ AstraLogic One-Stop Platform │ │ │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ Streamlit Dashboard with Agent Chat Interface │ │ │ │ "Plan a LEO-to-ground link with backup path" │ │ │ └─────────────────────────────────────────────────────────┘ │ │ ↓ │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ FastAPI Backend ← → OpenClaw Agent Decision Engine │ │ │ │ (Request routing, state management, result caching) │ │ │ └─────────────────────────────────────────────────────────┘ │ │ ↓ ↑ │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ OpenClaw Skills & MCP Servers │ │ │ │ • LinkBudgetSkill — analyze propagation & margins │ │ │ │ • FrequencyOptimizerSkill — select frequencies │ │ │ │ • AFSIMExecutor MCP — launch Warlock/Mystic headless │ │ │ └─────────────────────────────────────────────────────────┘ │ │ ↓ │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ AFSIM Simulation (Warlock/Mystic) │ │ │ │ Binary AER Output → pymystic Parser → CSV Results │ │ │ └─────────────────────────────────────────────────────────┘ │ │ ↓ │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ Dashboard Result Visualization │ │ │ │ Orbital plots, link metrics, performance summary │ │ │ └─────────────────────────────────────────────────────────┘ │ └────────────────────────────────────────────────────────────────┘ ``` --- ## Core Architecture AstraLogic implements an **Agent-Driven Simulation Pipeline**: | Component | Technology | Purpose | |-----------|-----------|---------| | **Web Dashboard** | Streamlit | User interface with Agent chat box for natural language requests | | **Backend Coordinator** | FastAPI | Routes agent decisions, manages simulation state, caches results | | **Decision Engine** | OpenClaw Agent | Reasons about orbital mechanics, link budgets, and resource optimization | | **Skill System** | Custom Skills | LinkBudgetAnalyzer, FrequencyOptimizer, ManeuverPlanner | | **Simulation Bridge** | MCP (Model Context Protocol) | Headless AFSIM execution, parameter templating, result collection | | **Data Parser** | pymystic + pandas | Binary AER → structured CSV for visualization | ### Workflow: End-to-End Simulation ``` User: "Plan a backup link for Sat-A when it loses contact with GS-1" ↓ (natural language → parsed request) OpenClaw Agent receives context: • Current orbital state (from last simulation) • Available frequency bands • Ground station locations • Link margin thresholds ↓ (reasoning loop) OpenClaw invokes Skills: 1. LinkBudgetSkill.analyze() → SNR for candidate frequencies 2. FrequencyOptimizer.select_channels() → optimal band selection 3. ManeuverPlanner.compute_handover() → timing & power levels ↓ (decision ready) AFSIMExecutor.run() via MCP: • Template WSL with agent-selected parameters • Launch Warlock/Mystic headless • Stream binary AER output • Parse results in real-time ↓ (results ready) Dashboard auto-updates with: • Orbital trajectory plots • Link margin time-series • Success/failure metrics • Agent reasoning summary ``` --- ## Key Features ### ✅ Active Development - 🤖 **Agent Chat Interface** — Natural language simulation requests directly in the dashboard - 🧠 **OpenClaw Decision Engine** — Intelligent agent for link planning, frequency optimization, and maneuver scheduling - 🎯 **Skill-based Architecture** — Extensible Skills for link-budget analysis, frequency selection, orbital mechanics - 🔌 **MCP Server Integration** — Headless AFSIM execution orchestrated via Model Context Protocol - 🔍 **Real-time AER Parser** — Binary simulation output ingestion with live CSV export - 📊 **Dashboard Visualization** — Plotly-based orbital plots, link metrics, and performance dashboards - ⚡ **FastAPI Backend** — REST API for state queries, result caching, and agent coordination - 🔄 **Automated Workflows** — Complete simulation from request to result without leaving the browser ### 📅 Future Enhancements - Advanced link-budget calculations (rain fade, fading margins) - Multi-scenario comparison & optimization loop - AFSIM direct launch button for advanced users - Expanded orbital mechanics solvers (perturbations, collision avoidance) --- ## Quick Start ### Prerequisites - Python ≥ 3.8 - AFSIM Warlock or Mystic installation (local or remote) - OpenClaw configuration (provided in setup) ### Installation ```bash git clone https://github.com/abydym/AstraLogic.git cd AstraLogic pip install -r requirements.txt ``` ### Running the Platform ```bash # Terminal 1: Start the backend API python backend.py # API available at http://localhost:8000/docs # Terminal 2: Launch the dashboard streamlit run app.py # Dashboard opens at http://localhost:8501 ``` ### Using the Agent Chat Interface Once the dashboard is live, use the **Agent Chat Box** to request simulations: **Example 1: Simple link analysis** ``` User: "Analyze the communication link between Sat-A and Ground Station-1 when Sat-A passes over the ground station at 14:30 UTC." OpenClaw will: 1. Retrieve orbital ephemerides 2. Compute elevation angle and slant range 3. Run LinkBudgetSkill to determine SNR margins 4. Launch AFSIM simulation via MCP 5. Display results in dashboard ``` **Example 2: Frequency optimization** ``` User: "Find the best frequency band for Sat-B to GS-2 link with at least 6 dB margin and no interference." OpenClaw will: 1. Analyze available frequency allocations 2. Check interference environment 3. Invoke FrequencyOptimizerSkill 4. Simulate multiple candidates in parallel 5. Recommend optimal band with reasoning ``` ### Custom Simulation (Advanced) For direct control, you can also invoke simulations programmatically: ```python from aermsg2dataframe import parse_aer_messages import pymystic import pandas as pd # Read existing AFSIM output with pymystic.Reader('scenarios/demo_output.aer') as reader: messages = list(reader) # Parse to DataFrames df_entity, df_orbital = parse_aer_messages(messages) # Analyze print(f"Simulation duration: {df_entity['simTime'].max()} seconds") print(f"Tracked platforms: {sorted(df_entity['platformIndex'].unique())}") # Compute inter-satellite distance sat1 = df_entity[df_entity['platformIndex'] == 1] sat2 = df_entity[df_entity['platformIndex'] == 2] if len(sat1) > 0 and len(sat2) > 0: distances = ((sat1[['x','y','z']].reset_index(drop=True) - sat2[['x','y','z']].reset_index(drop=True))**2).sum(axis=1)**0.5 / 1000 print(f"Min distance: {distances.min():.1f} km, Max: {distances.max():.1f} km") ``` --- ## Project Structure ``` AstraLogic/ ├── app.py # 🎨 Streamlit dashboard (main UI) ├── backend.py # ⚡ FastAPI server ├── aer_read.py # 📡 AER binary parser (pymystic wrapper) ├── aermsg2dataframe.py # 📊 Message → DataFrame converter ├── openclaw_agent.py # 🤖 LLM agent (planned) ├── afsim_mcp_server.py # 🔌 MCP server bridge (planned) ├── pymystic.py # 📦 Local pymystic library │ ├── core/ # Core utility modules │ ├── __init__.py │ └── afsim_bridge.py # AFSIM integration helpers │ ├── scenarios/ # AFSIM scenario files │ ├── demo_2sat_1gs.txt # 2 satellites + 1 ground station │ ├── demo_output.aer # Sample AFSIM output (binary) │ ├── demo_output.csv # Converted CSV (for reference) │ └── demo_output.evt # Event log │ ├── output/ # Generated data files │ ├── entity.csv # Parsed platform states │ └── orbital.csv # Parsed orbital elements │ ├── docs/ # Documentation ├── brand-design-md-main/ # Design system assets ├── README.md # This file ├── DESIGN.md # Architecture & design notes ├── requirements.txt # Dependencies ├── LICENSE # MIT License └── .env.example # Environment template (unused for now) ``` --- ## Roadmap | Milestone | Status | Target | Description | |-----------|--------|--------|-------------| | ✅ AER Parser | Complete | v0.1 | AFSIM binary deserialization (pymystic) | | ✅ Dashboard Core | Complete | v0.1 | Streamlit UI with orbital visualization | | ✅ FastAPI Backend | Complete | v0.1 | REST API for state management | | 🔨 Agent Chat Interface | In Progress | v0.2 | OpenClaw integration with natural language processing | | 🔨 LinkBudgetSkill | In Progress | v0.2 | Eb/N₀, EIRP, interference analysis | | 🔨 MCP Server Bridge | In Progress | v0.2 | Headless AFSIM execution & result streaming | | 📅 FrequencyOptimizerSkill | Planned | v0.3 | Automated frequency band selection | | 📅 Multi-scenario Optimization | Planned | v0.3 | Parallel simulation & comparative analysis | | 📅 Advanced Orbital Mechanics | Planned | v0.4 | Perturbations, collision avoidance, station-keeping | | 📅 Result Export & Reporting | Planned | v0.4 | PDF reports, data archiving, batch export | --- ## Contributing Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change. Please make sure to update tests as appropriate. --- ## License This project is licensed under the [MIT License](LICENSE) — see the LICENSE file for details. --- --- <a name="chinese"></a> ## 概述 **AstraLogic** 是一个一站式卫星通信仿真决策平台，在单一网页界面中实现了 AFSIM（高级仿真、集成与建模框架）与智能决策引擎的无缝融合。用户完全通过网页仪表盘进行交互——在 Agent 对话框中提交仿真需求，由 **OpenClaw** 智能体负责决策。OpenClaw 可以推理轨道力学、链路预算和通信策略，通过 **Skills** 和 **MCP 服务器**协调 AFSIM 的无头执行。仿真完成后，结果自动解析并在仪表盘中可视化展示，用户无需离开浏览器。系统架构包括： 1. **前端界面** — Streamlit 仪表盘，集成 Agent 对话框用于端到端仿真控制 2. **后端协调** — FastAPI 服务器，负责路由 Agent 决策和管理仿真状态 3. **决策引擎** — OpenClaw 智能体，推理轨道和通信优化 4. **技能系统** — 自定义 Skills 用于链路分析和频率优化 5. **仿真网桥** — MCP 协议支持无头 AFSIM 执行和结果导入 6. **数据层** — AER 解析器将 AFSIM 二进制输出转换为结构化 DataFrame ``` ┌────────────────────────────────────────────────────────────────┐ │ AstraLogic 一站式仿真平台 │ │ │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ Streamlit 仪表盘（含 Agent 对话框） │ │ │ │ "规划卫星A在失去地面站信号时的备用链路" │ │ │ └─────────────────────────────────────────────────────────┘ │ │ ↓ │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ FastAPI 后端 ← → OpenClaw Agent 决策引擎 │ │ │ │ （请求路由、状态管理、结果缓存） │ │ │ └─────────────────────────────────────────────────────────┘ │ │ ↓ ↑ │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ OpenClaw Skills 与 MCP 服务器 │ │ │ │ • LinkBudgetSkill — 分析传播与链路余量 │ │ │ │ • FrequencyOptimizerSkill — 频率选择 │ │ │ │ • AFSIMExecutor MCP — 无头启动 Warlock/Mystic │ │ │ └─────────────────────────────────────────────────────────┘ │ │ ↓ │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ AFSIM 仿真（Warlock/Mystic） │ │ │ │ 二进制 AER 输出 → pymystic 解析 → CSV 结果 │ │ │ └─────────────────────────────────────────────────────────┘ │ │ ↓ │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ 仪表盘结果展示 │ │ │ │ 轨道图、链路指标、性能总结 │ │ │ └─────────────────────────────────────────────────────────┘ │ └────────────────────────────────────────────────────────────────┘ ``` --- ## 核心架构 AstraLogic 实现了**智能体驱动的仿真管道**： | 组件 | 技术栈 | 职责 | |------|--------|------| | **网页仪表盘** | Streamlit | 用户界面与 Agent 对话框，用于自然语言仿真请求 | | **后端协调** | FastAPI | 路由 Agent 决策、管理仿真状态、缓存结果 | | **决策引擎** | OpenClaw Agent | 推理轨道力学、链路预算和资源优化 | | **技能系统** | 自定义 Skills | LinkBudgetAnalyzer、FrequencyOptimizer、ManeuverPlanner | | **仿真网桥** | MCP 协议 | 无头 AFSIM 执行、参数模板、结果收集 | | **数据解析** | pymystic + pandas | 二进制 AER → 结构化 CSV 用于可视化 | ### 工作流：端到端仿真 ``` 用户：「规划备用链路，当卫星A与地面站失联时」 ↓ （自然语言 → 解析请求） OpenClaw Agent 接收上下文： • 当前轨道状态（来自上次仿真） • 可用频段 • 地面站位置 • 链路余量阈值 ↓ （推理循环） OpenClaw 调用 Skills： 1. LinkBudgetSkill.analyze() → 候选频率的信噪比 2. FrequencyOptimizer.select_channels() → 最优频段选择 3. ManeuverPlanner.compute_handover() → 时序与功率参数 ↓ （决策就绪）通过 MCP 执行 AFSIMExecutor： • 用 Agent 选定的参数模板化 WSL • 无头启动 Warlock/Mystic • 实时流式处理二进制 AER 输出 • 在线解析结果 ↓ （结果就绪）仪表盘自动更新： • 轨道轨迹图 • 链路余量时间序列 • 成功/失败指标 • Agent 推理摘要 ``` --- ## 主要特性 ### ✅ 实际开发中 - 🤖 **Agent 对话框** — 在仪表盘中用自然语言提交仿真需求 - 🧠 **OpenClaw 决策引擎** — 智能体用于链路规划、频率优化和机动调度 - 🎯 **基于 Skill 的架构** — 可扩展的 Skills 用于链路分析、频率选择、轨道力学 - 🔌 **MCP 服务器集成** — 通过 Model Context Protocol 协调无头 AFSIM 执行 - 🔍 **实时 AER 解析器** — 二进制仿真输出摄取与实时 CSV 导出 - 📊 **仪表盘可视化** — 基于 Plotly 的轨道图、链路指标和性能仪表盘 - ⚡ **FastAPI 后端** — REST API 用于状态查询、结果缓存和 Agent 协调 - 🔄 **自动化工作流** — 从请求到结果的完整仿真，无需离开浏览器 ### 📅 后续扩展 - 高级链路预算计算（雨衰、衰减余量） - 多场景对比与优化循环 - AFSIM 直接启动按钮（供高级用户使用） - 扩展轨道力学求解器（扰动项、碰撞规避） --- ## 快速开始 ### 环境要求 - Python ≥ 3.8 - AFSIM Warlock 或 Mystic 安装（本地或远程） - OpenClaw 配置（安装包中提供） ### 安装 ```bash git clone https://github.com/abydym/AstraLogic.git cd AstraLogic pip install -r requirements.txt ``` ### 启动平台 ```bash # 终端 1：启动后端 API python backend.py # API 文档：http://localhost:8000/docs # 终端 2：启动仪表盘 streamlit run app.py # 仪表盘：http://localhost:8501 ``` ### 使用 Agent 对话框仪表盘启动后，使用 **Agent 对话框**提交仿真请求： **示例 1：简单链路分析** ``` 用户：「分析卫星A与地面站-1 在UTC 14:30 通过地面站上空时的通信链路」 OpenClaw 将： 1. 检索轨道星历 2. 计算仰角和斜距 3. 运行 LinkBudgetSkill 确定信噪比余量 4. 通过 MCP 启动 AFSIM 仿真 5. 在仪表盘中展示结果 ``` **示例 2：频率优化** ``` 用户：「为卫星B到地面站-2 的链路找到最优频段，至少6 dB余量且无干扰」 OpenClaw 将： 1. 分析可用频段分配 2. 检查干扰环境 3. 调用 FrequencyOptimizerSkill 4. 并行仿真多个候选频段 5. 推荐最优频段并说明理由 ``` ### 自定义仿真（高级用户）若需要直接控制，也可以编程方式调用仿真： ```python from aermsg2dataframe import parse_aer_messages import pymystic import pandas as pd # 读取现有 AFSIM 输出 with pymystic.Reader('scenarios/demo_output.aer') as reader: messages = list(reader) # 解析为 DataFrame df_entity, df_orbital = parse_aer_messages(messages) # 分析 print(f"仿真时长：{df_entity['simTime'].max()} 秒") print(f"追踪平台：{sorted(df_entity['platformIndex'].unique())}") # 计算星间距离 sat1 = df_entity[df_entity['platformIndex'] == 1] sat2 = df_entity[df_entity['platformIndex'] == 2] if len(sat1) > 0 and len(sat2) > 0: distances = ((sat1[['x','y','z']].reset_index(drop=True) - sat2[['x','y','z']].reset_index(drop=True))**2).sum(axis=1)**0.5 / 1000 print(f"最小距离：{distances.min():.1f} km，最大距离：{distances.max():.1f} km") ``` --- ## 项目结构 ``` AstraLogic/ ├── app.py # 🎨 Streamlit 仪表盘（主界面） ├── backend.py # ⚡ FastAPI 服务器 ├── aer_read.py # 📡 AER 二进制解析器（pymystic 封装） ├── aermsg2dataframe.py # 📊 消息 → DataFrame 转换器 ├── openclaw_agent.py # 🤖 LLM 智能体（筹划中） ├── afsim_mcp_server.py # 🔌 MCP 服务器网桥（筹划中） ├── pymystic.py # 📦 本地 pymystic 库 │ ├── core/ # 核心实用模块 │ ├── __init__.py │ └── afsim_bridge.py # AFSIM 集成助手 │ ├── scenarios/ # AFSIM 场景文件 │ ├── demo_2sat_1gs.txt # 2 颗卫星 + 1 个地面站 │ ├── demo_output.aer # 示例 AFSIM 输出（二进制） │ ├── demo_output.csv # 转换后的 CSV（参考） │ └── demo_output.evt # 事件日志 │ ├── output/ # 生成的数据文件 │ ├── entity.csv # 解析后的平台状态 │ └── orbital.csv # 解析后的轨道要素 │ ├── docs/ # 文档 ├── brand-design-md-main/ # 设计系统资源 ├── README.md # 本文件 ├── DESIGN.md # 架构与设计笔记 ├── requirements.txt # 依赖清单 ├── LICENSE # MIT 许可证 └── .env.example # 环境变量模板（暂未使用） ``` --- ## 路线图 | 里程碑 | 状态 | 目标版本 | 描述 | |--------|------|--------|------| | ✅ AER 解析器 | 已完成 | v0.1 | AFSIM 二进制反序列化（pymystic） | | ✅ 仪表盘核心 | 已完成 | v0.1 | Streamlit UI 与轨道可视化 | | ✅ FastAPI 后端 | 已完成 | v0.1 | 状态管理 REST API | | 🔨 Agent 对话框 | 进行中 | v0.2 | OpenClaw 集成与自然语言处理 | | 🔨 LinkBudgetSkill | 进行中 | v0.2 | Eb/N₀、EIRP、干扰分析 | | 🔨 MCP 服务器网桥 | 进行中 | v0.2 | 无头 AFSIM 执行与结果流式处理 | | 📅 FrequencyOptimizerSkill | 筹划中 | v0.3 | 自动频段选择 | | 📅 多场景优化 | 筹划中 | v0.3 | 并行仿真与对比分析 | | 📅 高级轨道力学 | 筹划中 | v0.4 | 扰动项、碰撞规避、台站保持 | | 📅 结果导出与报告 | 筹划中 | v0.4 | PDF 报告、数据存档、批量导出 | --- ## 贡献指南欢迎提交 Pull Request。对于重大变更，请先提交 Issue 进行讨论。请确保同步更新相关测试。 --- ## 许可证本项目采用 [MIT 许可证](LICENSE)，详见 LICENSE 文件。

Carregando o README do repositório…

Skills neste repositório

Engenheiros eletricistas

1 categorias ocupacionais · 100% classificado

skill

ocupação

descrição

atualizado

afsim-project

Engenheiros eletricistas

AFSIM 2.9.0 仿真技能：创建、编译、调试、可视化和分析仿真场景。工作流：写 .txt → mission.exe 编译 → 修复错误 → mystic.exe 可视化 → Python 分析。

Idioma do texto original: chinês

2026-06-11

Mostrando 1 de 1 skills coletadas.