llama-cpp
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues detected. The skill serves as a reference for deploying and optimizing llama.cpp.
- [EXTERNAL_DOWNLOADS]: Fetches source code from the official ggerganov/llama.cpp GitHub repository and models from HuggingFace. These are trusted sources for this ecosystem.
- [INDIRECT_PROMPT_INJECTION]: The skill documents interfaces for processing external data via CLI prompts and API messages in SKILL.md. While this is an ingestion surface for untrusted data, it represents the core intended functionality of the tool and uses standard delimiters in examples. Capability inventory includes local inference execution and model file management.
- [COMMAND_EXECUTION]: Provides standard build (make) and execution commands for the llama.cpp binaries. These commands are consistent with the tool's intended use and do not involve privilege escalation or obfuscated payloads.
Audit Metadata