ClawHub
Back to skill

Security audit

Chain of Density

Security checks for vulnerabilities and agentic risk

Overview

This skill provides summarization instructions plus a local text-counting helper, with no evidence of hidden data access, persistence, or destructive behavior.

Installers should expect an instruction-only summarization workflow that may send the provided source text through a serial subagent process, plus a local word-count helper. Avoid using it for legal, compliance, or other text where exact wording must be preserved, as the skill itself also warns.

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 (1)

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The declared purpose centers on generating or refining summaries using Chain-of-Density. The supplied code does not summarize, transform, or densify text at all. It only counts words, characters, and UTF-8 bytes, and exposes those counts via a small command-line interface. While such metrics could support a summarization workflow, they are merely auxiliary and do not match the claimed primary behavior. Therefore the description materially misrepresents what the code chunk actually does.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.