company-research

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) due to its core function of processing untrusted external data.
  • Ingestion points: External data enters the agent context via Exa search results including company metadata, news articles, and social media posts (tweets).
  • Boundary markers: The skill instructions do not define clear delimiters or specific instructions for the agent to ignore potentially malicious commands embedded in search results.
  • Capability inventory: The skill possesses significant capabilities including spawning task agents and performing browser automation via a Chrome fallback mechanism.
  • Sanitization: No sanitization, filtering, or validation logic is specified for the external content before it is processed by the LLM intelligence.
  • Mitigating Factor: The 'Token Isolation' instruction acts as a partial architectural defense by confining untrusted data processing to isolated sub-agents, preventing the main context from direct exposure.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 04:51 AM