agent-pagerank-analyzer
Pass
Audited by Gen Agent Trust Hub on Mar 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the vendor's mcp__flow-nexus__sandbox_execute tool to run Python scripts for distributed graph processing. Evidence is found in the 'Distributed Graph Processing' section of SKILL.md.
- [PROMPT_INJECTION]: The ingestion of complex graph data from external sources creates a surface for indirect prompt injection. 1. Ingestion points: Adjacency matrices and influence vectors in SKILL.md. 2. Boundary markers: Not present in the provided examples. 3. Capability inventory: Sandboxed Python execution and specialized matrix solvers. 4. Sanitization: No input validation or sanitization logic is explicitly defined.
Audit Metadata