Back to skill

Security audit

OpenClaw Agent Orchestrator

Security checks for vulnerabilities and agentic risk

Overview

ClawScan verdict: benign, with medium confidence; scanners found a configuration-reading concern, while the other assessed areas were rated okay.

The verdict is benign, but not risk-free. Before installing, consider that the configuration loader may read credential-bearing sections even though the scanner reported no credential output or external transmission. Coverage was partial: several references in SKILL.md could not be resolved, and the QR image was not analyzed by static scanners.

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
  • 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 configuration loader reads all configuration sections despite permission being restricted to selected orchestration fields.

Medium
Category
Excessive Agency
Confidence
91% confidence
Finding

The manifest explicitly limits configuration access to agent IDs and specified agentToAgent/session settings, excluding credentials and other configuration sections. L54–L58 instead read the complete file and deserialize every section into memory, including credential-bearing sections if present; filtering the fields used later does not enforce the declared read boundary. No external transmission or credential output is shown.

Content

No source excerpt is available for this finding.