embodied-navigation
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external text, URLs, and conversation history, which creates a potential entry point for indirect prompt injection.
- Ingestion points: The core method involves passing text and URLs to MCP tools like optimize_text_structure and generate_topical_clusters.
- Boundary markers: The instructions do not define specific delimiters or guardrail prompts to isolate processed data from the agent's core logic.
- Capability inventory: The skill utilizes InfraNodus MCP tools for situational mapping and diagnostic purposes.
- Sanitization: There is no documentation of internal sanitization or validation of the ingested content prior to analysis.
Audit Metadata