wzry-build-advisor
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use the
greptool to search for hero and equipment information within the local reference files (references/heroes.md,references/equipment.md). This is a standard and legitimate use of agent capabilities for data retrieval within the skill's own package. - [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect injection because it processes untrusted user input (hero names and game state) and uses it to perform tool-based searches.
- Ingestion points: User-provided strings for heroes and equipment are used as search terms in
SKILL.md. - Boundary markers: The instructions do not specify sanitization or delimiters for the user-provided search terms.
- Capability inventory: The agent is authorized to use
grepto read local reference files. - Sanitization: There is no explicit instruction to sanitize or validate the hero names before they are used in search queries.
Audit Metadata