niceboard

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

• Third-party content exposure detected (high risk: 0.80). The skill instructs the agent to use the Membrane CLI to connect to Niceboard and run actions (e.g., membrane action run ...) that retrieve Niceboard data such as customer feedback (user-generated content) and return JSON outputs the agent is expected to read and act on, exposing it to untrusted third‑party content from Niceboard as shown in SKILL.md.