Back to skill

Security audit

Openclaw Extension Update

Security checks for vulnerabilities and agentic risk

Overview

The skill is rated suspicious with medium confidence because scanners report update actions without authorization and deletion of duplicate skill directories beyond the stated update and inventory scope.

Review the reported update and directory-removal behavior before installing. The verdict is suspicious, not malicious; Skillspector's analysis is partial because several referenced paths were not found in the bundled artifact, so the evidence is incomplete.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (1)

T09 · Insecure Skill Coding Practices

Warning
Location
SKILL.md
Finding

Inspection-only requests lead to force updates without update authorization

Content
View full analysis

Impact

An inspection request can install unapproved skill revisions and overwrite local modifications. Backups support recovery but do not prevent unauthorized replacement.

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 (1)

The skill directs deletion of duplicate skill directories, exceeding the manifest's update-and-inventory scope.

Medium
Category
Excessive Agency
Confidence
94% confidence
Finding

L087–L088 instruct the agent to remove skill copies using gio trash and then verify the retained copy. The manifest describes updating extensions and inventorying them, not deleting installations; removing duplicate directories is a separate state-changing operation rather than a necessary part of checking or applying updates. Although recoverable, this can change which installed copy is available.

Content

No source excerpt is available for this finding.