ClawHub
Back to skill

Security audit

OpenClaw Automation Architecture

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation-only automation design skill whose guidance is broadly purpose-aligned and does not include hidden code, credential handling, or exfiltration behavior.

This skill is reasonable to install if you want OpenClaw automation design help. Review its recommendations before enabling cron jobs, spawned agents, external messages, or workflows that touch business-critical data, because the skill can advise on persistent and recurring automation patterns.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Vague Triggers

Medium
Confidence
96% confidence
Finding
The trigger metadata is very broad, including common phrases like 'save time', 'pipeline', 'reminder', and 'automate this', which can cause this skill to activate for many ordinary requests outside its intended scope. Over-broad activation increases the chance that the agent applies automation-specific guidance inappropriately, potentially leading to unintended task execution design, overcollection of context, or risky delegation recommendations.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
- Prefer **small reliable systems** over giant brittle flows.
- Separate **trigger**, **execution**, **state**, **delivery**, and **recovery**.
- Pick the cheapest primitive that can do the job well.
- Do not ask the user to choose among primitives unless the trade-off materially affects behavior, cost, or reliability.

## Quick Selection
Confidence
80% confidence
Finding
Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
- Do not add an external platform if `cron` + tools + scripts already solve it.
- Do not build giant all-in-one jobs when two small jobs with a file handoff are clearer.
- Do not rely on repeated polling if eventing or longer waits work.
- Do not send external messages without approval when approval is required.
- Do not put business logic only in your head; store it in files, prompts, scripts, or config.
- Do not make every failure page the user. Some failures should log quietly and retry later.
Confidence
75% confidence
Finding
Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Static analysis

No suspicious patterns detected.