web-search-advanced-personal-site

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [No Code] (SAFE): The skill consists only of markdown instructions and does not include any scripts, executables, or package manifests.
  • [Prompt Injection] (SAFE): The use of 'Critical' markers is strictly for directing the agent toward specific tool parameters and isolation practices; no attempts to bypass safety filters or extract system prompts were detected.
  • [Indirect Prompt Injection] (LOW): The skill enables the ingestion of untrusted content via web search. Evidence Chain: 1. Ingestion points: Results from web_search_advanced_exa. 2. Boundary markers: None specified in the prompt. 3. Capability inventory: The skill only summarizes data; it has no file-write, network-send (other than the search tool itself), or command-execution capabilities. 4. Sanitization: Not specified.
  • [Data Exfiltration] (SAFE): There are no hardcoded credentials or attempts to access sensitive local file paths. All network operations are confined to the intended web search tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:24 PM