Back to skill

Security audit

Gateway Recovery

Security checks for vulnerabilities and agentic risk

Overview

The skill is transparent about sending a recovery message automatically, but it relies on persistent startup/heartbeat behavior and gives too little control over where and when the message is sent.

Review before installing. This skill should only be used if you want automatic recovery messages after gateway restarts and you can verify who receives them, how often they can be sent, how the LaunchAgent is installed or removed, and where logs are written.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • 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)

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
91% confidence
Finding

This skill is designed to autonomously decide when to send a recovery message based on a restart/heartbeat condition, with no human confirmation step. Autonomous message sending can be abused or misfire, causing unintended communications, disclosure of system state, or message spam if the trigger is manipulated.

Content

Scanner excerpt · SKILL.md (reported line 4)May include surrounding context.

md
# Gateway Recovery Notification Skill

## Purpose
Automatically sends a notification message when the gateway restarts and recovers.

## How It Works

Session Persistence

Medium
Category
Rogue Agent
Confidence
75% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 15)May include surrounding context.

md
## Configuration

- **Flag file**: `~/.openclaw/recovery_flag`
- **LaunchAgent**: `ai.openclaw.gateway-startup-hook.plist`
- **Log file**: `/Users/kazuni/.openclaw/workspace/memory/startup.log`

## Usage

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill explicitly states it runs automatically during heartbeat checks and sends a notification without manual intervention, but it does not present this as a clear user-facing warning or consent boundary. Silent outbound messaging is security-relevant because it can surprise users, leak presence/state information, and normalize autonomous communications without explicit approval.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 25)May include surrounding context.

md
## Testing

To test manually:
1. Create flag: `touch ~/.openclaw/recovery_flag`
2. Secretary will detect it on next heartbeat
3. Send notification message
4. Clear flag automatically