Company Analysis

Pass

Audited by Gen Agent Trust Hub on Mar 3, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill's behavior is entirely consistent with its purpose as a research and enrichment tool. It follows a rigorous methodology for analyzing public business data and does not attempt any unauthorized file access or command execution.
  • [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests and processes untrusted data from external websites during its research phases.
  • Ingestion points: Public web content from sources such as LinkedIn, G2, and company blogs is fetched via the web_search capability as described in the Analysis Methodology.
  • Boundary markers: The instructions do not define explicit delimiters to isolate fetched research data from the system's reasoning process.
  • Capability inventory: The skill is restricted to web_search and structured data generation; it does not have the ability to write files, execute scripts, or perform network requests to non-whitelisted domains.
  • Sanitization: No explicit sanitization or filtering of external content is specified before the data is synthesized into insights.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 3, 2026, 11:37 AM