phenomenon-based-unit-anchor
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, such as prompt injection, data exfiltration, or obfuscation, were detected. The skill is limited to text generation for curriculum design and does not request any dangerous permissions or tools.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an ingestion surface for untrusted data as it interpolates user-provided text into the system prompt without explicit delimiters or sanitization. However, this surface is not exploitable as the skill does not utilize any tools with network, file, or system access. 1. Ingestion points: Input fields defined in the input_schema (phenomenon, subjects_involved, etc.) in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: Limited to text generation; no subprocess, network, or file system capabilities are enabled. 4. Sanitization: Absent, relying on the model's internal safety guardrails.
Audit Metadata