The Agent Skills Directory

Remote Dependency (HIGH): The skill instructs users to add https://rube.app/mcp as an MCP server. This domain is not among the trusted sources (e.g., Anthropic, OpenAI, Microsoft). This establishes a critical dependency on unverified infrastructure that controls all tool logic.
Indirect Prompt Injection (HIGH): The skill exhibits a high-risk attack surface for processing malicious external content.
Ingestion points: Processes raw data from Foursquare API responses through RUBE_SEARCH_TOOLS and tool execution outputs.
Boundary markers: There are no instructions or delimiters provided to help the agent distinguish between Foursquare data and potentially malicious embedded instructions.
Capability inventory: The skill allows for writing and executing actions on a user's Foursquare account, including bulk operations via RUBE_REMOTE_WORKBENCH.
Sanitization: No sanitization or validation of the external data is performed before it influences the agent's next steps.
Dynamic Tool Execution (MEDIUM): The workflow relies on RUBE_MULTI_EXECUTE_TOOL using schemas fetched at runtime from a remote server. This allows the remote provider to change execution patterns or input requirements dynamically, which could be used to facilitate unauthorized operations if the remote server is compromised or malicious.

foursquare-automation