100Critical
The skill executes user-provided repository scripts, enabling
Behavior Analysis
Description verified — behavior matches claim
Claims to do
WordPress Router: Use this skill at the start of most WordPress tasks to:
Actually does
The skill executes a Node.js script (`detect_wp_project.mjs`) to analyze the current working directory (repo root). It reads the script's output to classify the WordPress project type, available tooling, and testing frameworks. It then uses this information, along with a local decision tree file (`decision-tree.md`), to determine the appropriate subsequent workflow.
Skill Info
Namewordpress/agent-skills/wordpress-router
Registrygithub
Version2ec2cd8e9da7
PURLpkg:github/wordpress/agent-skills@2ec2cd8e9da7?skill=wordpress-router
Stars1,417
Installs1,400
Source
SHA-25638f4dc335679...
SHA-1edf0940915a4...
MD5d02a9179108c...
TLSHca418639bb48...
Size2,227 bytes
Install
This skill has been flagged as potentially malicious. Review the findings below before installing.
Skills.shDocs
npx skills add https://github.com/wordpress/agent-skillsAssessments (1)
AI Agent Traps ↗Execution of arbitrary repository scriptscriticalcommand execution
The agent is explicitly instructed to execute `lint`, `test`, or `build` commands found within the user's repository. This allows an attacker to embed malicious scripts in their repository, which the agent will then execute.
100% confidenceline 40
Run the repo’s lint/test/build commands that the triage output recommends (if available).
Broad system command execution capabilitieshighcommand execution
The skill explicitly states compatibility with `bash`, `node`, and `WP-CLI`, indicating the agent has broad capabilities to execute system commands and scripts. While necessary for its function, this presents a significant attack surface if not properly sandboxed.
100% confidenceline 4
compatibility: "Filesystem-based agent with bash + node. Some workflows require WP-CLI."
Filesystem access and reconnaissancelowreconnaissance
The agent is designed to operate on the 'Repo root (current working directory)' and inspect various configuration files, confirming its ability to read local filesystem content. This capability, while functional, could be leveraged for reconnaissance.
100% confidenceline 24
Inputs required: - Repo root (current working directory). ... inspect: - root composer.json, package.json, style.css, block.json, theme.json, wp-content/.
Badge
Markdown
[](https://mondoo.com/ai-agent-security/skills/github/wordpress/agent-skills/wordpress-router)HTML
<a href="https://mondoo.com/ai-agent-security/skills/github/wordpress/agent-skills/wordpress-router"><img src="https://mondoo.com/ai-agent-security/api/badge/github/wordpress/agent-skills/wordpress-router.svg" alt="Mondoo Skill Check" /></a>Image URL
https://mondoo.com/ai-agent-security/api/badge/github/wordpress/agent-skills/wordpress-router.svgSecure your AI agents
Skills can read files, run commands, and access credentials. Mondoo helps organizations manage the security risks of AI agent skills across their entire fleet.
- Continuous scanning of skills across all registries
- Policy enforcement before skills reach your agents
- Integration with your existing security workflow