sdpm-style

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No security issues detected. The skill facilitates interaction with the sdpm MCP server to generate presentation style guides. It correctly delegates logic to the server and enforces a specific initialization sequence.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and follow instructions returned from the sdpm server at runtime. While this allows external content to influence the agent's behavior, it is the intended primary function for providing dynamic, server-side style definitions. The skill includes clear boundaries by instructing the agent to wait for these instructions before proceeding.
  • [COMMAND_EXECUTION]: The skill requires calling a specific tool (start_presentation) on the sdpm server. This is a legitimate use of the MCP (Model Context Protocol) framework and does not constitute unauthorized command execution or privilege escalation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 09:53 AM
Security Audit — agent-trust-hub — sdpm-style