flow-simplify
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [NO_CODE]: The skill consists entirely of natural language instructions and guidelines without any associated scripts, executables, or automated command sequences.\n- [INDIRECT_PROMPT_INJECTION]: The skill instructions involve analyzing and modifying external codebases, which represents a potential attack surface if the code being processed contains malicious instructions disguised as logic or comments.\n
- Ingestion points: The process involves analyzing source code, imports, dynamic lookups, registries, and configuration files (SKILL.md).\n
- Boundary markers: None specified to differentiate between analyzed code data and agent instructions.\n
- Capability inventory: The skill implies the ability to read and potentially modify (delete) files within a repository.\n
- Sanitization: No specific sanitization or validation of the input code content is described.
Audit Metadata