separate-stems-api

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (medium risk: 0.30). The skill contains deceptive mismatch steering, advertising an audio stem separation API while instructing the model to perform image generation and handle image prompts.

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

• Third-party content exposure detected (low risk: 0.10). The required workflow fetches and ingests an external documentation file from a URL (https://www.pixazo.ai/models/separate-stems-api.md), which constitutes low-risk ingestion of arbitrary web content.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs the agent to fetch external instruction/documentation files from https://www.pixazo.ai/models/separate-stems-api.md, constituting an unverifiable runtime external dependency from an unestablished publisher.