multi-model-meta-analysis

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION] (LOW): The skill is susceptible to indirect prompt injection because its core workflow involves processing untrusted data (outputs from other AI models) and using that data to guide tool execution.
  • Ingestion points: SKILL.md instructs the agent to parse and extract claims from 'feedback/analysis from multiple LLMs' provided by the user.
  • Boundary markers: Absent. The skill provides no instructions to wrap untrusted data in delimiters or warnings to ignore embedded instructions within the analyzed content.
  • Capability inventory: The skill explicitly directs the agent to use high-privilege filesystem tools, specifically Grep, Glob, and Read, to verify claims across the codebase.
  • Sanitization: Absent. There is no requirement for the agent to sanitize the extracted claims before using them to drive file-searching logic, which could allow a malicious claim to trick the agent into accessing sensitive files (e.g., .env or config files) and including their contents in the final 'Synthesized Assessment'.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:32 PM