behavioral-modes

Fail

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: HIGHPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION] (HIGH): The skill establishes a high-risk surface for Indirect Prompt Injection by design.
  • Ingestion points: The REVIEW, DEBUG, and EXPLORE modes are designed to read and analyze untrusted external content, including third-party code, architecture descriptions, and error logs.
  • Boundary markers: Absent; the instructions lack delimiters or specific directives for the agent to ignore or isolate instructions found within the data it is processing.
  • Capability inventory: The IMPLEMENT mode provides the agent with the authority to write production-ready code, and the allowed-tools (Read, Glob, Grep) grant broad visibility into the file system. This allows an injection to manipulate file-write operations.
  • Sanitization: Absent; no logic is provided to sanitize or escape content before the agent incorporates it into its reasoning or output.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 17, 2026, 07:59 AM