ClawHub
Back to skill

Security audit

Department Manager

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed local task-tracking helper for AI departments and does not show evidence of hidden execution, credential access, networking, or destructive behavior.

Install this if you want a local department/task tracker for OpenClaw workflows. Be aware that task descriptions and completion outputs are stored on disk in ~/.openclaw/department-manager/departments.json, so avoid entering secrets or sensitive business data unless you are comfortable with that local persistence.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
89% confidence
Finding
The skill exposes commands that create, assign, complete, and remove department state, and explicitly writes persistent data to ~/.openclaw/department-manager/departments.json, yet it declares no permissions or allowed-tools scope. That mismatch can cause the host agent to invoke a file-writing skill without clear least-privilege boundaries, increasing the chance of unintended state changes or abuse through broad tool access.

Vague Triggers

Medium
Confidence
89% confidence
Finding
The invocation description is very broad: 'Manage a team of AI sub-agents' and 'essential for autonomous agents running businesses or complex projects' could match many generic planning, delegation, or coordination prompts. This overbroad trigger surface may cause the skill to be selected in contexts where multi-agent task delegation, persistent task tracking, or file-backed state changes were not intended, amplifying operational and security risk.

Missing User Warnings

Low
Confidence
93% confidence
Finding
This code creates a directory under the user's home folder and uses it to persist department/task data, but the file does not include a docstring, comment, or other disclosure explaining that state will be stored on disk in ~/.openclaw/department-manager. For a code file, persistent file writes are in scope when they lack any visible warning or explanatory disclosure.

Missing User Warnings

Low
Confidence
95% confidence
Finding
The save_data function writes all departments, tasks, and outputs to departments.json, but there is no comment, help text, or warning indicating that user-supplied task content and outputs will be persisted locally. The operation is visible in code but not disclosed in a user-facing way beyond generic success messages.

Static analysis

No suspicious patterns detected.