The Agent Skills Directory

[External Downloads] (LOW): The skill requires the configuration of an external MCP server at https://rube.app/mcp. This source is not on the predefined trusted list, meaning the tools and logic it provides are unverifiable.
[Indirect Prompt Injection] (LOW): The workflow relies on dynamically fetching tool schemas and execution plans from the remote server via RUBE_SEARCH_TOOLS and then executing them. This architecture is susceptible to tool output poisoning if the external server returns malicious instructions.
Ingestion points: RUBE_SEARCH_TOOLS API responses.
Boundary markers: Absent; the skill instructs the agent to use discovered schemas directly.
Capability inventory: RUBE_MULTI_EXECUTE_TOOL and RUBE_REMOTE_WORKBENCH provide broad operational capabilities, including remote execution of tasks.
Sanitization: None; the instructions emphasize exact field name compliance from the untrusted search results without validation.

beaconstac-automation