respira-builder-edits
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill describes a legitimate administrative workflow for WordPress content management. It includes technical details for specific builder versions and emphasizes safety through snapshots and verification steps.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it involves reading and processing data from external WordPress pages.
- Ingestion points: Data is ingested via the
respira_read_page,respira_find_element, andrespira_get_page_outlinetools in SKILL.md. - Boundary markers: The workflow instructs the agent to use specific element selectors (text, CSS class, or element ID) and outlines, which helps bound the processing of external content to targeted nodes.
- Capability inventory: The skill uses tools with write capabilities, including
respira_update_elementandrespira_batch_update, to modify site content. - Sanitization: The documentation references the WordPress
ksessystem, noting that while standard filtering applies, specific bypasses are maintained for functional code-heavy modules to preserve existing scripts and styles during edits.
Audit Metadata