dependency-mapping
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided project or system documents to generate dependency maps. This creates a surface where instructions hidden within these documents could potentially influence the agent's logic.
- Ingestion points: SKILL.md Phase 1 accepts external subject descriptions and references.
- Boundary markers: Uses a structured interview mode but lacks explicit delimiters for document content.
- Capability inventory: The skill generates reports and triggers diagram rendering tools (Phase 9).
- Sanitization: Normalizes input into noun-based nodes and unique IDs.
- [COMMAND_EXECUTION]: The skill utilizes the
mmdc(Mermaid CLI) tool to render diagrams and export them as PNG files, which involves executing external commands on the system.
Audit Metadata