Back to skill

Security audit

OpenClaw Release Announcement

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marked this skill suspicious because scanners report that it updates the local OpenClaw installation across release channels, a state-changing capability outside its announcement-focused purpose.

Consider the reported OpenClaw updates and their lack of interactive confirmation before installing. The scan is partial: two local references could not be resolved, so evidence is incomplete. A separate scan completed without findings, but that does not rule out risks.

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)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The declared purpose is release-announcement preparation or posting, but the code instead performs a local software update. Displaying the version could support release validation, but changing the installed software is an undeclared state-changing capability. The supplied code does not draft or post an announcement.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

Checking the installed version could support release validation, but executing an update changes the local installation rather than drafting or posting an announcement. This is a materially different capability not disclosed in the declared purpose.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

Checking the installed version could support release validation, but performing an automatically confirmed software update is a distinct, state-changing capability not disclosed by the announcement-focused description. The supplied chunk does not draft or post an announcement.

Content

No source excerpt is available for this finding.