NYC

wp-abilities-api

Pass

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: SAFE
Full Analysis
  • Prompt Injection (SAFE): No instructions were found that attempt to bypass AI safety filters, override system prompts, or extract internal configuration.
  • Data Exposure & Exfiltration (SAFE): The skill does not access sensitive local files (e.g., credentials, SSH keys) or perform network requests to untrusted external domains.
  • Obfuscation (SAFE): The content is clear and uses standard Markdown/YAML formatting with no hidden characters or encoded payloads.
  • Unverifiable Dependencies & Remote Code Execution (SAFE): The skill references standard development tools (WP-CLI, Node, Bash) and a specific library (@wordpress/abilities). There are no commands that download and execute arbitrary scripts from the internet (e.g., curl | bash).
  • Privilege Escalation (SAFE): No commands involving 'sudo', 'chmod', or administrative modifications to the host system were identified.
  • Persistence Mechanisms (SAFE): The skill does not attempt to modify shell profiles, cron jobs, or system services to maintain unauthorized access.
  • Indirect Prompt Injection (LOW): As a coding assistant skill that reads and analyzes repository content, there is a theoretical surface for instructions embedded in analyzed files to influence the agent. However, this is a standard risk for all filesystem-based agents and no specific vulnerabilities were introduced in this skill's logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 19, 2026, 06:37 PM