onethread

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 Onethread and run actions (e.g., "membrane action list" / "membrane action run") to fetch and interact with Onethread data—a customer-support platform containing user-generated, untrusted messages—which the agent is expected to read and act on per the SKILL.md workflow.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill requires fetching and running the Membrane CLI (e.g., "npm install -g @membranehq/cli@latest" and "npx @membranehq/cli@latest"), which downloads and executes remote code at runtime and is required for the skill to operate, so it is a runtime external dependency that can execute code.