dineway-seo-schema

Pass

Audited by Gen Agent Trust Hub on Aug 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions are focused on logical data validation and template generation. It does not contain any executable code, obfuscation, or remote resource downloads.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes existing schema and facts from external web pages (Ingestion: Workflow steps 1 and 2). It mitigates injection risks through explicit validation (Sanitization: Workflow step 6) and operational boundaries (Boundary: 'Do not use local provider credentials', 'Do not use local fetch/parse/render commands'). The capability is limited to text-based recommendations (Capability: Audit and generate schema).
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The instructions include specific boundaries against using local provider credentials, preventing the exposure of sensitive secrets.
  • [UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill explicitly forbids the use of local fetch, parse, or render commands, preventing unauthorized system activity.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 13, 2026, 10:29 AM
Security Audit — agent-trust-hub — dineway-seo-schema