The Agent Skills Directory

[Prompt Injection] (SAFE): No patterns of instruction overrides, jailbreaks, or system prompt extraction were found in the skill content.
[Data Exposure & Exfiltration] (SAFE): No hardcoded credentials or sensitive local file path access detected. Authentication is handled via a managed connection flow (RUBE_MANAGE_CONNECTIONS).
[Indirect Prompt Injection] (LOW): The skill has a surface for indirect injection by processing external tool schemas.
Ingestion points: Tool definitions and schemas returned by the RUBE_SEARCH_TOOLS endpoint.
Boundary markers: Absent. The skill does not explicitly instruct the agent to use delimiters or ignore embedded instructions within search results.
Capability inventory: The skill can execute actions via RUBE_MULTI_EXECUTE_TOOL based on ingested schemas.
Sanitization: Not mentioned. The agent is expected to trust the output of the Rube MCP server.
[External Downloads] (SAFE): The skill references an external MCP endpoint (https://rube.app/mcp) and documentation (composio.dev) required for its stated purpose. No execution of untrusted remote scripts (e.g., curl|bash) was found.
[Obfuscation] (SAFE): No encoded strings, zero-width characters, or homoglyphs were detected.

apex27-automation