local-llm-expert
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides educational content and configuration guidance for local LLM inference engines. All documented commands (e.g.,
ollama run,llama.cpparguments) are standard industry usage. - [SAFE]: No external network dependencies, remote code execution, or data exfiltration patterns were found. The skill emphasizes local privacy and offline execution.
- [SAFE]: Prompting instructions are strictly related to LLM chat templates (ChatML, Zephyr) and do not contain injection or bypass attempts.
Audit Metadata