skills/sales-skills/sales/sales-g2/Gen Agent Trust Hub

sales-g2

Warn

Audited by Gen Agent Trust Hub on Apr 20, 2026

Risk Level: MEDIUMPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill contains deceptive content regarding industry acquisitions, claiming G2 acquired Capterra, Software Advice, and GetApp in February 2026. This misinformation can lead users and the agent to make incorrect assumptions about platform capabilities and the competitive landscape.
  • [PROMPT_INJECTION]: The skill implements a persistent 'learnings' mechanism in references/learnings.md that stores knowledge from user interactions. This creates an indirect prompt injection surface because the file is read in subsequent sessions without sanitization or boundary markers. Ingestion points: User-provided questions and tips are used to populate the references/learnings.md file. Boundary markers: None identified; entries are read directly as context. Capability inventory: The agent can append to local files and route tasks to other specialized skills. Sanitization: There is no logic to validate or sanitize user-provided tips before they are persisted.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 20, 2026, 06:35 PM
Security Audit — agent-trust-hub — sales-g2