perplexity
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of untrusted external content by performing web searches and conversational queries.
- Ingestion points: Data enters the agent's context through the outputs of
mcp__perplexity__perplexity_searchandmcp__perplexity__perplexity_askinSKILL.md. - Capability inventory: The agent can perform network-based web searches and retrieve summarized content from the internet.
- Boundary markers: The skill includes instructions to limit result sizes (
max_results: 3,max_tokens_per_page: 512), which restricts the volume of data but does not explicitly frame the content to prevent the agent from following instructions embedded in search results. - Sanitization: No explicit sanitization, filtering, or validation logic is defined in the instructions to handle potentially malicious content within retrieved web pages.
Audit Metadata