Back to skill

Security audit

OpenClaw Subagent Toolset

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed prompt guide for choosing sub-agent tool subsets and contains no executable code, persistence, exfiltration, or hidden install behavior.

Before installing, treat this as a prompt-template aid rather than an enforced sandbox. Use Coding mode only for user-directed coding tasks, and resolve the Secretary glob_search contradiction if strict tool boundaries matter.

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
  • 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 (3)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The README presents the skill description in Chinese only, with no indication that users may choose another language or that the locale restriction is required for a region-specific purpose. This can violate language/locale policy because it imposes a specific language on users without opt-in.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The Coding subset grants write and command-execution capabilities and the skill later states that tool subsets are only soft constraints, meaning safety depends on prompt compliance rather than enforcement. That combination can lead to unintended file modification or unsafe command use by a spawned sub-agent, especially if the parent task prompt is ambiguous or adversarial.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The Secretary section says the sub-agent may use glob_search in its allowed tools list, but the Forbidden line also explicitly forbids glob_search. This is an active contradiction in the skill documentation about intended allowed behavior.

Content

No source excerpt is available for this finding.