bitwarden-automation

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 call RUBE_SEARCH_TOOLS against an external MCP (e.g., the rube.app/mcp/Composio tool registry and associated toolkit schemas like composio.dev/toolkits/bitwarden) and to read returned tool schemas and execution plans, which are third‑party, runtime-provided data the agent must interpret—exposing it to untrusted external content that could carry indirect prompt injections.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill instructs adding and using https://rube.app/mcp as the MCP server at runtime (via RUBE_SEARCH_TOOLS and RUBE_MULTI_EXECUTE_TOOL) to fetch tool schemas, recommended execution plans, and execute tools, so remote content from that endpoint can directly control agent behavior and run code.