ClawHub
Back to skill

Security audit

Hash Toolkit

Security checks for vulnerabilities and agentic risk

Overview

This skill is a small local hashing utility with no signs of hidden access, persistence, or data exfiltration.

Install only if a basic local hashing helper is sufficient. Do not rely on its perceptual hash for strong image similarity, near-duplicate detection, or security-sensitive deduplication without replacing it with a real perceptual hashing algorithm.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The function is labeled as a perceptual hash but actually computes a normalized-text SHA256 substring, which is not resilient to semantically similar but syntactically different inputs the way true perceptual hashing is expected to be. In a deduplication or similarity workflow, this can cause false trust in similarity detection, allowing near-duplicate or modified content to evade detection or causing incorrect matching decisions.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Low
Category
Not specified by scanner
Confidence
92% confidence
Finding

The manifest description at L005 limits the advertised scope to MD5, SHA256, and perceptual hashing for deduplication. The code in generateMultipleHashes additionally computes SHA1 and a truncated SHA512 value, expanding the skill's behavior beyond what is described.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.