Search Documentation

Pass

Audited by Gen Agent Trust Hub on Mar 3, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface because it processes untrusted user queries and external data to synthesize answers. In Phase 4 (Synthesis), the LLM reads article content retrieved from an external store, which could contain malicious instructions designed to manipulate the agent's behavior.\n
  • Ingestion points: The query input parameter in SKILL.md and documentation content retrieved via the search_docs action.\n
  • Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands used when the agent processes retrieved documentation content.\n
  • Capability inventory: The skill utilizes the crm capability and executes searches via the search_docs internal tool.\n
  • Sanitization: No sanitization is performed on documentation text before it is presented to the LLM for synthesis, although keyword extraction is used for the search phase.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 3, 2026, 11:38 AM