Security audit
Security checks for vulnerabilities and agentic risk
This skill is a planning checkpoint that reads project context and pauses for confirmation before work, with no evidence of hidden execution, persistence, or data exfiltration.
Install this if you want the agent to pause and produce a reconnaissance report before larger or riskier coding work. Expect it to read relevant repository files and slow down some workflows because its trigger conditions are intentionally broad.
The skill says it activates on 'ANY' of several broad conditions such as touching more than one file, involving data processing, or being described as 'big', 'long', or 'risky'. These conditions lack clear scope boundaries or exclusions, so the skill could trigger for a very large portion of normal work rather than a narrowly defined context.
Skill grants unrestricted tool access without appropriate constraints. An agent with unfettered tool access can perform arbitrary actions including file modification, network requests, and code execution.
### 5. Stop
Do not write any code. Do not run any command beyond inspection. Do not make a plan beyond the report. Wait for the user.
The purpose of this skill is to prevent wasted effort, not to complete work. Stopping is the job.
No suspicious patterns detected.