The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill uses WebFetch to ingest documentation from developers.webflow.com. While this is an official source, skills that process external content possess an inherent attack surface for indirect instructions.
Ingestion points: WebFetch tool used on developers.webflow.com documentation endpoints.
Boundary markers: The skill does not define specific delimiters or "ignore instructions" markers for the fetched documentation.
Capability inventory: The skill has file-system access (Read, Glob, Grep, Write, Edit) and network access (WebFetch).
Sanitization: No explicit sanitization or validation of the fetched external documentation is described.
[SAFE]: The skill implements strong credential safety protocols, explicitly instructing the agent to never log, persist, or echo authentication tokens (bearer tokens, site tokens, or workspace tokens) in any output, command, or report.
[SAFE]: State-changing actions (create, update, delete, publish, etc.) are restricted by mandatory approval boundaries, ensuring the agent requires explicit user consent and provides rollback plans before proceeding with mutations.
[SAFE]: All external resources and documentation links are hosted on official Webflow domains, which are well-known and trusted developer service endpoints.

webflow-sdk-patterns