prompt-template-builder

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of documentation and code snippets providing a framework for prompt engineering. The provided TypeScript and Markdown examples demonstrate standard software development practices for building and testing AI-driven applications.- [INDIRECT_PROMPT_INJECTION]: The skill defines templates that interpolate external data (source code for review) into prompts via placeholders like ${variables.code}.
  • Ingestion points: The variables.code property in the codeReviewPrompt function in SKILL.md accepts external content.
  • Boundary markers: The template uses triple backticks (```) to delimit the external code.
  • Capability inventory: The skill documentation describes a workflow using a generic llm() function to process the resulting prompts, but the skill itself does not possess direct file-write or network-access capabilities.
  • Sanitization: While the input is not sanitized, the documentation strongly emphasizes the use of strict output contracts (Zod schema validation) and few-shot examples to ensure the model adheres to a specific JSON structure regardless of the input content, which is a key defense against indirect injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 09:30 AM
Security Audit — agent-trust-hub — prompt-template-builder