Language: Mandarin (中文)

As the RISC-V ecosystem rapidly evolves, how do we avoid having AI Agents and current Large Language Models (LLMs) "hallucinate", and how do we put AI Agents to use in the daily workflow of RISC-V?

In this 1-hour virtual meetup, we will dive deep into the practical synergy of the RVA23 Profile and introduce several skills designed for reading and cross-referencing RISC-V ISA specs based on UnifiedDB.

We will discuss how to build a practical skill using UnifiedDB, equipping AI Agents with precise execution capabilities.

We will also explore skills tailored for open-source software coding practices, especially focusing on the Linux kernel. Essentially, this approach combines RISC-V spec and kernel skills directly into the host's daily workflow.

Furthermore, we will share real-world experiences, pitfalls, and best practices in applying these AI skills to open-source software workflows.

Speakers (Hosted by)

Guodong Xu (Open-Source RISC-V Software Ecosystem Practitioner, Director of SW Engineering, RISCstar)

Focuses on low-level software architecture evolution, Linux Kernel cross-architecture migration, and the integration of AI Toolchains/Agent Skills.

Jian Zhang (Open-Source RISC-V CPU Product Manager) Deeply engaged in RISC-V hardware ecosystem, specification analysis, and hardware-software co-design.