kernel-hardening
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the analysis of external system data, creating an ingestion surface where instructions embedded in kernel config files could potentially influence agent behavior.
- Ingestion points: Identified in SKILL.md as kernel configuration files (.config/Kconfig), boot parameters, and system runtime settings.
- Boundary markers: The instructions do not specify the use of delimiters or markers to isolate the audited data from the agent's core instructions.
- Capability inventory: The skill explicitly restricts operations to read-only analysis and states it does not modify kernel configurations or execute system commands, which limits the potential impact of injection.
- Sanitization: The instructions specify that no user-provided values are used in commands or file paths.
Audit Metadata