wp-file-upload-security

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill provides guidance on processing untrusted user data through various ingestion points and identifies the associated security surface.
  • Ingestion points: The skill handles data from $_FILES arrays, multipart form submissions, REST API endpoints, and remote file sideloading via download_url (found in SKILL.md).
  • Boundary markers: It promotes the use of WordPress nonces via check_admin_referer and authorization checks using current_user_can to establish trust boundaries (found in SKILL.md).
  • Capability inventory: The skill utilizes powerful WordPress filesystem and media APIs, including media_handle_upload, media_handle_sideload, unzip_file, and wp_delete_attachment (found in SKILL.md).
  • Sanitization: It mandates the use of wp_check_filetype_and_ext for byte-level verification, sanitize_file_name for filename safety, and explicit MIME allowlists to filter malicious content (found in SKILL.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 07:07 PM
Security Audit — agent-trust-hub — wp-file-upload-security