analytics-engineering

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.90). The "Companion check" section instructs the agent to run local shell commands (ls) and offer installs (npx) to enumerate/modify the host environment, which is outside the Analytics Engineering skill's stated purpose and thus constitutes a hidden/deceptive operational instruction.

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s required runtime workflow is driven by the user’s free-text prompts (e.g., when the user asks about dbt setup, metrics, or self-serve patterns), and that text is directly ingested to produce the response—there is no indication of a trusted, non-user-authored-only ingestion path.