seo-video
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes local Python scripts to validate audit data and build reports. Evidence:
python "<suite-root>/scripts/validate_seo_findings.py"inSKILL.mdandpython "<suite-root>/scripts/manage_markdown_site.py"inreferences/markdown-delivery-contract.md. - [DYNAMIC_EXECUTION]: Script paths are dynamically resolved based on environment variables or configuration files. Evidence:
<suite-root>is determined from.seo-suite-runtime.jsonor${CLAUDE_PLUGIN_ROOT}. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted external content from video watch pages and transcripts.
- Ingestion points: External watch pages (raw/rendered), transcripts, and captions.
- Boundary markers: Absent.
- Capability inventory: File system writing and Python script execution.
- Sanitization: Absent.
Audit Metadata