pre-modification-check
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources (Repowise analysis tool outputs) which directly influences the agent's decision-making regarding code modifications, refactoring, and deletions.
- Ingestion points: Data enters the context via the
get_risk,get_context, andget_healthtool outputs inSKILL.md. - Boundary markers: The instructions lack specific delimiters or warnings to ignore potentially malicious instructions embedded within the codebase analysis data.
- Capability inventory: The agent is explicitly instructed to use this data to perform file modifications and deletions.
- Sanitization: No sanitization or validation logic is specified for the data returned by the tools.
Audit Metadata