cms-editorial-calendar

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The instructions provide guidance on workflow execution and output formatting. There are no attempts to bypass safety filters, extract system prompts, or assume unrestricted roles.
  • [DATA_EXPOSURE]: The skill is restricted to read-only access and only interacts with CMS data through specialized tools. It does not access credentials, sensitive local files, or communicate with external, untrusted domains.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external CMS content, which is a potential surface for indirect injection. However, the skill only formats this data into display tables and does not use the content to trigger further automated actions or execute commands, maintaining a safe posture.
  • [COMMAND_EXECUTION]: No arbitrary shell commands or subprocess executions are identified. All tool interactions are limited to the provided CMS-specific data retrieval functions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 02:57 PM
Security Audit — agent-trust-hub — cms-editorial-calendar