The Agent Skills Directory

[DATA_EXFILTRATION]: The skill transmits telemetry data, including usage duration, success status, and widget counts, to the vendor's tracking endpoint at https://www.respira.press/api/skills/track-usage. This is standard telemetry for skill performance monitoring.
[CREDENTIALS_UNSAFE]: To access design data, the skill requests a Figma personal access token from the user. This is a legitimate requirement for the Figma REST API and is handled as dynamic user input rather than being hardcoded.
[EXTERNAL_DOWNLOADS]: Downloads image assets from Figma's servers to the local WordPress media library via the respira_sideload_image tool. Figma is a well-known service and the operation is central to the skill's purpose.
[PROMPT_INJECTION]: The skill ingests untrusted data from Figma layers and text nodes, representing an indirect prompt injection surface.
Ingestion points: Figma node tree structure and content via REST API.
Boundary markers: The execution workflow mandates a 'Phase 2: Build plan' where the agent must present the conversion details for human approval before proceeding.
Capability inventory: Can create and modify WordPress pages using tools like respira_build_page and respira_create_page_duplicate.
Sanitization: The skill instructions specify that the Respira plugin validates and normalizes the generated layout data before it is applied to the database.

figma-to-elementor