The Agent Skills Directory

[SAFE]: The skill instructions define a specific workflow for transforming HTML into WordPress block structures using an MCP server, maintaining a clear separation between the agent's logic and the site's execution environment.
[SAFE]: No evidence of prompt injection, unauthorized data access, or malicious command execution was found. The documentation explicitly confirms the removal of dangerous capabilities like eval in the associated WordPress plugin.
[SAFE]: The skill implements content handling practices, including an allowlist for permitted HTML tags and automated stripping of sensitive tags like <script> to prevent cross-site scripting (XSS) in the generated blocks.
[SAFE]: External resource interactions, such as fetching media, are restricted to specific MCP tools that operate within the WordPress security framework, avoiding raw shell-based network operations.

canai-blocks