v2.0.1a Release Notes
[Important: Support Policy & Terms of Use]
This project is a general-purpose platform designed to integrate various industrial equipment and legacy systems with AI agents. To maintain stable operation and high-quality technical support, official support for this firmware is strictly limited to the following hardware:
- Industrial & Legacy Systems: Legacy PCs via RS-232C (CP/M, MSX, Apple II, etc.) and industrial serial instruments.
- Specific Telescope Systems: DOG NS-5000 / NS-3000 and the OnStepNinja Series.
For generic OnStep clones or other custom hardware not listed above, the operation of the advanced Streamable HTTP and SSE protocols implemented in this firmware is NOT guaranteed. To ensure the sustainable development and integrity of this project, we kindly ask that you refrain from using this firmware on unsupported or unauthorized systems.
[Overview]
AiBridgeMCP: A Universal Bridge Connecting Industrial Equipment and AI
- Instant AI Systems: Transform existing industrial or vintage equipment into AI-driven systems instantly. By providing the device manual to the AI, you can enable natural language hardware control without writing a single line of code.
- Advanced Protocol Implementation: Featuring a custom-engineered implementation of the Streamable HTTP protocol on the resource-constrained ESP32. This version is fully compliant with the latest Model Context Protocol (MCP) specifications, ensuring robust and seamless connectivity with AI clients like Claude Desktop.
[Key Changes in v2.0.1a]
- Protocol Overhaul: Transitioned from SSE to Streamable HTTP. Now supports full bi-directional communication based on the latest MCP standards.
- Stability Improvements: Multiple fixes to enhance serial bridge connectivity and the operational stability of the ESP32 core.
- MCP Optimization: Optimized performance and compatibility for the latest MCP Inspector and Claude Desktop environments.
[Use Cases & Applications]
- Industrial & Legacy Integration: Enables real-time, natural language dialogue between modern AI and legacy systems (CP/M, MSX, etc.) or specialized industrial measurement tools via RS-232C.
- Manual-Based Hardware Control: Simply feed a hardware device's technical manual to the AI to generate immediate control capabilities. Operate complex machinery using natural language without memorizing command sets.
- Example Application — Telescope Control: Direct control of LX200-compatible mounts via DOG NS-5000 / OnStepNinja. Build a sophisticated "AI Astronomy Assistant" capable of querying coordinates and performing autonomous slews.