review-ml-textbook

Pass

Audited by Gen Agent Trust Hub on Aug 23, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted external data such as online textbooks and web URLs without explicit sanitization or boundary markers. 1. Ingestion points: The workflow maps artifacts from external URLs and chapter files provided by the user. 2. Boundary markers: The instructions do not define clear delimiters or instructions to ignore embedded prompts within the textbooks. 3. Capability inventory: The skill utilizes subagents to process different chapters, which can be influenced by malicious content. 4. Sanitization: No filtering or validation of the ingested content is performed.
  • [EXTERNAL_DOWNLOADS]: The skill fetches data from external, user-provided web URLs as part of its core textbook review function.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 23, 2026, 10:38 PM
Security Audit — agent-trust-hub — review-ml-textbook