Back to skill

Security audit

CS Reply Pack

Security checks for vulnerabilities and agentic risk

Overview

ClawScan’s verdict is benign with high confidence: the scanners report no AIG findings, while the noted risks concern installation commands rather than the stated customer-service reply-template purpose.

The verdict is benign, but installation is not risk-free: if you follow the documented installation commands, an unpinned external package may be downloaded and run. Consider verifying the installer and skill source before installing.

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
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • 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 (3)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The supplied chunk performs installation rather than generating or applying customer-service reply rules. Remote package installation and external CLI execution are not described capabilities. Although the filename suggests release documentation and the package name matches the advertised skill, neither establishes that the installed content implements the declared functionality.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx skills'.

Medium
Category
MCP Rug Pull
Confidence
91% confidence
Finding

The documented command can download and execute the unpinned npm package skills, making installation dependent on whichever version resolves at execution time. Although this is an installer rather than an MCP server and no malicious behavior is shown, a compromised release could execute with the installing user's permissions; the risk arises during installation, not ordinary customer-service template use.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx skills'.

Medium
Category
MCP Rug Pull
Confidence
91% confidence
Finding

The release notes recommend executing npx skills without a version pin, allowing the externally resolved package version to change independently of this release. This is not an MCP server reference, but it is a supply-chain exposure if a user follows the installation instructions; nothing here demonstrates actual compromise.

Content

No source excerpt is available for this finding.