course-storytelling
Pass
Audited by Gen Agent Trust Hub on Aug 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill incorporates a local utility script,
scripts/verify.sh, which is used to perform heuristic analysis on lesson files to ensure they meet the framework's quality standards. The script is read-only and performs local pattern matching using standard system utilities. - [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it is designed to ingest and process untrusted user data such as lecture transcripts, slide decks, and course outlines.
- Ingestion points: Untrusted data is ingested from user-provided files stored in
02-DOCS/raw/teaching/. - Boundary markers: No specific delimiters or "ignore previous instruction" warnings are employed when processing the ingested content.
- Capability inventory: The skill uses the agent to analyze and rewrite text, and executes the local
scripts/verify.shutility. - Sanitization: No explicit sanitization or validation of the ingested content is performed prior to the agent processing it.
Audit Metadata