social-proof-injection

Pass

Audited by Gen Agent Trust Hub on Apr 3, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by instructing the agent to incorporate external, potentially untrusted data (prospect inputs and testimonial database content) into its conversational output.
  • Ingestion points: Prospect attributes such as stated_needs, pain_points, and objections_raised, as well as the content field from the social proof library (SKILL.md).
  • Boundary markers: Absent; the provided Python snippets (e.g., format_testimonial) use direct string interpolation without delimiters or instructions to ignore embedded commands within the data.
  • Capability inventory: The skill focuses on message construction and formatting logic for automated sales interactions.
  • Sanitization: No sanitization, escaping, or validation of the external content is implemented or recommended in the provided implementation examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 3, 2026, 05:46 PM