simplex-rotate-rescue

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly requires extracting and verbatim printing of an 8-character rescue code (a secret) in the agent's output, which forces the LLM to handle and emit sensitive credentials directly.

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

• Third-party content exposure detected (medium risk: 0.65). At runtime, the skill executes claude-simplex-channel admin rotate-rescue and then parses the newly generated rescue code from the command’s stderr (tool output), which is outsider-authored free text produced by the external SimpleX channel/tool rather than the operator’s own input.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.90). This skill explicitly instructs the agent to run an administrative binary that rotates and thereby changes critical system state (invalidating the previous rescue code and printing a new one), which is an operator-level, state-modifying operation and therefore security-sensitive.