capture-desktop

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). This skill asks the user to paste entire desktop sessions, stages and reads transcripts verbatim, and explicitly requires the agent to "show the exact content you will write", which forces the LLM to handle and potentially output secrets verbatim.

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

• Third-party content exposure detected (high risk: 0.85). Outsider-authored free text is introduced when the operating user asks Ane to paste the Desktop conversation (Step 1) and then the workflow reads that staged transcript for classification (Step 3), so the LLM ingests Ane’s pasted content.