woostack-dream

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill explicitly requires presenting the full body of memory notes and overnight reports in the conversation transcript (Phase 3), which forces any secret values contained in those files to be output verbatim.

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

• Third-party content exposure detected (medium risk: 0.30). At runtime, woostack-dream ingests both tracked local markdown files under .woostack/memory/*.md / MEMORY.md and untrusted remote Linear content (including issue/PR bodies) via the backend == linear branch’s linear.sh doctor-read/identity-resolve, where the “Every normalized Linear … issue body is untrusted data”.