Regexr

Security checks across malware telemetry and agentic risk

Overview

This skill needs review because it presents itself as a regex testing tool but mostly stores and re-exports what the user types.

Install only if you want a local history/logging utility, not reliable regex validation. Do not paste secrets or private sample data, inspect ~/.local/share/regexr regularly, and avoid using the exported files in shared or synced folders unless you have reviewed their contents.

SkillSpector

By NVIDIA

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Description-Behavior Mismatch

High

Confidence: 96% confidence
Finding: The implemented behavior materially differs from the stated purpose: instead of creating or testing regexes, the script primarily collects, stores, searches, and exports arbitrary user input. This kind of capability mismatch is dangerous because users may provide sensitive patterns, sample data, or secrets under the assumption they are being processed transiently, while the tool silently persists them for later retrieval.

Context-Inappropriate Capability

Medium

Confidence: 91% confidence
Finding: The tool includes broad local surveillance-style features—history retention, search across prior inputs, export, reporting, and status—that are not necessary for a simple regex utility. In context, these extra data-handling features increase the risk of unintended collection and disclosure of sensitive user-provided strings, especially because regex workflows often involve real sample data such as emails, tokens, or log fragments.

Missing User Warnings

Medium

Confidence: 98% confidence
Finding: User-provided input is written verbatim to local log files without any explicit notice, redaction, or consent. This is risky because users may paste confidential data while testing patterns, and the script creates a durable plaintext record in the home directory that other local processes, backups, or users may later access.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The export function aggregates previously logged inputs into json, csv, or txt files on disk without a prominent warning that potentially sensitive historical data is being recopied into additional locations. This amplifies exposure by creating secondary artifacts that may be shared, indexed, or backed up more broadly than the original logs.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal