automation-flow-generate

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The skill mandates scanning local sfdx project metadata and repeatedly sending that metadata to a specified external MCP tool (execute_metadata_action) without user confirmation or pause, which creates a high risk of automated, unauthorized data exfiltration and abuse.

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s runtime workflow ingests outsider-authored free text directly via the required MCP calls where the user’s natural language request (userPrompt) is passed into fetchGroundedObjectMetadata and flowElementSelection before flowElementGeneration.