Security audit
Security checks for vulnerabilities and agentic risk
This is a straightforward text-processing skill with ordinary file read/write behavior and no evidence of hidden execution, persistence, exfiltration, or privilege escalation.
Before installing, treat it as a local text-file utility: use explicit input and output paths, avoid overwriting important files, and be careful when extracting emails or phone numbers from sensitive documents. I found no evidence that it sends data elsewhere or modifies the system persistently.
The skill advertises file-reading and file-writing workflows through its commands, but it does not declare any explicit tool scope, permissions, or allowed-tools boundaries. That creates an authorization and least-privilege gap: an agent may infer broad file access is acceptable and perform unintended reads or overwrites without clear policy constraints.
The description uses broad activation wording like 'use when users need to process, format, clean, or analyze text content,' which can cause the skill to trigger for a very wide set of routine user requests. Overbroad invocation increases the chance the agent applies file-modifying or extraction behaviors in situations where the user did not intend to delegate those actions to this skill.
The overview presents the skill as a general-purpose, everyday text toolkit without clear invocation limits or safety boundaries. In context, that ambiguity is more concerning because the documented features include regex replacement, file output, and data extraction, which can change user data or surface sensitive content if invoked too freely.
The usage examples show output-writing and text-modifying operations, but the markdown does not warn that these commands may overwrite existing files or alter data. Without an explicit caution, users or agents may run destructive operations on important files, leading to accidental data loss or corruption.
The skill explicitly supports extracting URLs, emails, and phone numbers from text, but it provides no privacy notice or handling guidance for potentially sensitive data. In a broad-use skill, that omission can normalize bulk extraction of personal information without prompting the user to consider consent, minimization, or downstream exposure.
No suspicious patterns detected.
