ClawHub
Back to skill

Security audit

everything-search

Security checks across malware telemetry and agentic risk

Overview

This is a coherent local file search skill, but it can automatically run bundled Windows search binaries that broadly index and expose local file information.

Install only if you are comfortable with a bundled Everything instance searching broad local file metadata and possibly remaining running after use. Prefer an already trusted local Everything installation, and avoid using content searches on sensitive locations unless you intend those contents to be searched.

SkillSpector

Vulnerability Patterns
  • Memory PoisoningPersistent Context Injection, Context Window Stuffing, Memory Manipulation
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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 (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
86% confidence
Finding
The skill instructs the agent to invoke a shell command (`python scripts/search.py ...`) and references bundled executables, but the metadata does not declare any corresponding permissions. Undeclared shell capability is a security issue because it bypasses permission transparency and can enable execution of local binaries or scripts with user-controlled arguments, increasing the risk of command injection or unintended process launching.

Context Window Stuffing

Medium
Category
Memory Poisoning
Content
�{u�
PH��H;�|�H�\$0H�l$8H�t$@H�|$HH�� A\���������Hc�H�ƓH�H�
�H�%(P��������H�\$H�t$H�|$ATH�� HcپH�=�u�����N
�<����������H��H�L�%e�I�<�t���y�(�p���H��H��u������
3��X�
��I�<�u.��H�����uH������������
3��I�
Confidence
81% confidence
Finding
The embedded ASCII string 'PADDINGXXPADDING...' is unusual in a production DLL and can indicate intentionally inserted filler data to distort scanners, confuse downstream parsers, or hide meaningful content among noise. While not directly exploitable by itself, in the context of an agent skill supply chain it is a red flag because opaque binary dependencies with anomalous filler reduce auditability and may conceal malicious modifications.

VirusTotal

61/61 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.