ClawHub
Back to skill

Security audit

Resume After Gateway Restart

Security checks for vulnerabilities and agentic risk

Overview

This skill transparently schedules a short-lived restart/resume workflow for OpenClaw gateway restarts, with no evidence of hidden or unrelated behavior.

Install this only if you are comfortable with a skill that can schedule one-shot OpenClaw cron agent turns and restart the OpenClaw gateway after a short delay. Before invoking it, verify the session key and resume note are correct, because the follow-up agent turn will act on memory/post-restart-task.md and reply to the captured chat route.

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 (2)

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

The declared description is about reliable post-restart resumption of agent work, including durable scheduling of a future agent turn and replying back to the original chat after restart. The supplied code only schedules a delayed restart of the gateway service via systemd-run. While this may be a supporting piece of a restart workflow, by itself it does not implement the core declared behavior of resuming work after restart, preserving model/thinking/session characteristics, or ensuring post-restart message delivery. Therefore the code's actual behavior is materially narrower and different from the declared purpose.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
85% confidence
Finding

This shell script deletes existing cron jobs and creates new ones, which changes system/task scheduling state and can affect automated behavior after restart. Although the script comments describe what it does, there is no confirmation prompt or prominent user-facing warning before these state-changing operations execute.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.