The Agent Skills Directory

[SAFE]: The skill serves as a guide for integrating Etermin automation via the Rube MCP server. The workflows described—tool discovery, connection verification, and tool execution—are standard for the Composio ecosystem.
[PROMPT_INJECTION]: The skill was evaluated for Indirect Prompt Injection (Category 8) vulnerability surfaces.
Ingestion points: The agent ingests external data in the form of tool schemas and recommended execution plans from the RUBE_SEARCH_TOOLS output.
Boundary markers: No explicit delimiters or safety warnings regarding the content of the search results are present in the instructions.
Capability inventory: The skill utilizes RUBE_MULTI_EXECUTE_TOOL and RUBE_REMOTE_WORKBENCH to execute commands based on the ingested metadata.
Sanitization: There is no evidence of sanitization or validation of the schemas retrieved from the remote MCP server.
Assessment: While the skill creates a surface where the agent acts on externally provided metadata, this behavior is the core intended functionality of the Rube/Composio automation pattern and does not constitute a malicious finding in this context.

etermin-automation