Skills
skills/davila7/claude-code-templates/ai-wrapper-product/Gen Agent Trust Hub

ai-wrapper-product

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • PROMPT_INJECTION (LOW): The skill's implementation patterns for AI wrappers create a surface for indirect prompt injection.
  • Evidence:
  • Ingestion points: The userInput variable in the generateContent function and properties of the input object in the promptTemplates (SKILL.md).
  • Boundary markers: The provided code examples lack explicit delimiters (such as XML tags or unique section headers) to isolate untrusted user data from system instructions within the interpolated prompt string.
  • Capability inventory: The skill facilitates interaction with external LLM APIs (Anthropic SDK).
  • Sanitization: The implementation includes a basic character length check (userInput.length > 5000), but does not provide logic to sanitize or escape input against malicious prompt injection payloads.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 04:56 PM