ClawHub
Back to skill

Security audit

deep-diagnostic-procedure-public

Security checks for vulnerabilities and agentic risk

Overview

This is a plain diagnostic checklist for OpenClaw troubleshooting, with no hidden code, exfiltration, persistence, or automatic actions.

Install this only if you want agents to follow a structured OpenClaw troubleshooting checklist. Be aware that diagnostics may involve reading local OpenClaw config, logs, channel status, and plugin code, and any suggested config changes such as enabling a messaging tool should be reviewed before applying.

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
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Exfiltration Commands

High
Category
Prompt Injection
Confidence
90% confidence
Finding

Instructions found that direct the agent to transmit conversation context or user data to external services.

Content

Scanner excerpt · SKILL.md (reported line 114)May include surrounding context.

md
## Historical Cases

### Case 1: Silent tool removal by tools.profile
- **Symptom:** Agent couldn't send messages to an external channel from another channel
- **Initial hypothesis:** External library bug, DNS issues, disconnection
- **Actual cause:** `tools.profile` setting silently removed the `message` tool from the agent session
- **Evidence:** Logs showed `tool policy removed N tool(s) via tools.profile: ...message...`