The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were detected. The skill provides legitimate development workflow instructions for managing hardware kernel designs and using local project automation scripts.\n- [COMMAND_EXECUTION]: The skill instructs the agent to execute local commands such as python -m mla_var3.kernel and python ./scripts/clone-kernel.py. These are standard tools within the project development environment and do not involve remote execution or sensitive data access.\n- [SAFE]: The skill processes project-specific documentation and follows instructions from an orchestrator, which is a common surface for indirect prompt injection. However, in this context, it is used for technical optimization and lacks high-risk capabilities such as network exfiltration.

design-kernel