d2l-discussion-responder
Pass
Audited by Gen Agent Trust Hub on Nov 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted student discussion posts, which is a potential surface for indirect prompt injection. However, the instructions provide strong mitigations by explicitly requiring that student content be treated only as data to be answered and forbidding the execution of any scripts or instructions embedded in the posts.
- [DATA_EXFILTRATION]: The skill handles sensitive student identifiers and thread content while performing necessary network operations to the LMS. It contains detailed safeguards to ensure student names, grades, and posts remain within the authorized session and are never stored in the skill's source control, shared memory, or public logs.
Audit Metadata