respira-workflow
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates interaction with live WordPress sites, which involves reading untrusted content that could contain malicious instructions.
- Ingestion points: Content is read from WordPress pages and posts via tools like respira_get_builder_info.
- Boundary markers: The skill does not explicitly define prompt boundary markers in this file, though it refers to a separate wordpress-editing-safety skill.
- Capability inventory: The skill can write to WordPress sites, create draft duplicates, and report issues via respira_report_issue.
- Sanitization: Employs validator_warnings and render_validator_pass to check if changes rendered correctly, and requires human approval for sensitive actions via respira_approval_required.
- [EXTERNAL_DOWNLOADS]: The skill references an external GitHub repository (github.com/respira-press/agent-skills-wordpress) for additional functionality. This repository belongs to the official vendor.
Audit Metadata