enterprise-togaf-zachman
Warn
Audited by Gen Agent Trust Hub on Jul 26, 2026
Risk Level: MEDIUMPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill contains extreme amounts of repetitive filler text across eight reference files (e.g., architecture-patterns.md, security-best-practices.md, testing-strategies.md). Each file contains 150 sections of near-identical text, which is deceptive and appears designed to hinder security audits or manipulate model attention and context limits.
- [PROMPT_INJECTION]: The skill ingests untrusted user architectural requirements without implementing sanitization or boundary markers, exposing the agent to indirect prompt injection. Ingestion points: User architectural descriptions and triggers in SKILL.md. Boundary markers: Absent; no delimiters are used to isolate user-provided data. Capability inventory: Generates structured enterprise architecture artifacts and viewpoints. Sanitization: Absent; no input validation or filtering is performed.
Audit Metadata