input-capture

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). This skill documents an input-capture API that legitimately enables interception of raw keystrokes (including passwords/tokens) and forwarding them via events/WebSocket — a high-abuse capability (keylogging/data exfiltration) though not inherently a backdoor in itself.

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

• Third-party content exposure detected (medium risk: 0.65). Outsider free text can enter the LLM context via the “Text Input” workflow: captured keystrokes from the human are received as input WebSocket events (raw/parsed) and the skill instructs buffering printable characters into a free-form string, which is then typically used in subsequent LLM reasoning.