kf-g-agent-review-adversarial-evaluation
Pass
Audited by Gen Agent Trust Hub on Aug 15, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze user-provided artifacts such as designs, proposals, and plans, creating a surface for indirect prompt injection. Instructions hidden within these artifacts could attempt to manipulate the agent's logic during the review process. * Ingestion points: User-provided plans, designs, and proposals specified in SKILL.md. * Boundary markers: Absent; the instructions do not define delimiters to isolate untrusted content. * Capability inventory: The skill possesses no capabilities for file writing, network communication, or command execution. * Sanitization: No input sanitization or validation is implemented.
- [NO_CODE]: The skill consists entirely of Markdown instructions and contains no executable scripts (Python, Node.js), shell commands, or remote resource fetching.
Audit Metadata