check-deliverable

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). This skill tells a spawned agent to read and grep a user file and then include the "full verbatim qa_block" in its output, which forces verbatim inclusion of any secrets present in the file and therefore risks secret exfiltration.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). In the /check-deliverable workflow, the runtime LLM work is a single-file review where it “reads and Grep[s] the file directly” at the user-supplied <file> path, so the outsider-authored free text (the deliverable content) is ingested.