plan-page

Pass

Audited by Gen Agent Trust Hub on Oct 12, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill integrates with an external Model Context Protocol (MCP) server located at https://mcp.trylexsis.com/mcp. This server provides necessary tools for the skill's operation, including product catalog access, review retrieval, and template management via the lexsis_catalog, lexsis_template_library, and lexsis_capture namespaces.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external sources, such as customer reviews, campaign messages, and web research, which presents a surface for indirect prompt injection.
  • Ingestion points: The skill reads data from products.md, fetches live customer reviews via catalog tools, processes user-supplied campaign materials, and performs authoritative web research for factual verification.
  • Boundary markers: The instructions emphasize distinguishing between "observed statements and inferred motivations" and require recording the source and date for all researched facts to maintain context integrity.
  • Capability inventory: The skill possesses the capability to query catalog data, retrieve customer reviews, search template libraries, and configure interactive Quiz components. These components use a JSON-based schema for logic and results.
  • Sanitization: A structured "Proof ledger" and "Claim gate" system is used to cross-reference every number, quote, and claim against verified sources before they are included in the final output.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 12, 2026, 05:56 AM