explore-graph
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: Executes the vendor's nmem CLI tool to retrieve system status and expand graph nodes.
- [SAFE]: Implements security guardrails by explicitly instructing the agent to never include API keys or credentials in generated URLs and requiring session verification before opening links.
- [INDIRECT_PROMPT_INJECTION]: Processes external memory data for visualization. Ingestion points: Memory IDs and CLI output (SKILL.md). Boundary markers: Depth and limit constraints applied to expansion. Capability inventory: MCP tool calls, CLI execution, and URL generation. Sanitization: URL-encoding of query parameters.
Audit Metadata