ClawHub
Back to skill

Security audit

Workspace Review

Security checks across malware telemetry and agentic risk

Overview

This workspace-audit skill appears coherent and purpose-aligned, but users should be aware it can inspect local workspace files and discusses durable memory practices.

Install only if you want an agent to review your OpenClaw workspace. Keep API keys, passwords, and personal data out of markdown and memory files, and invoke the skill explicitly for workspace maintenance rather than generic file review.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Rogue AgentSelf-Modification, Session Persistence
  • 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
91% confidence
Finding
The skill description is broad enough to match common requests like 'review', 'audit files', or 'check structure', which increases the chance of unintended invocation. Over-broad triggers can cause the agent to run this skill in contexts the user did not intend, potentially leading to unnecessary filesystem inspection or disclosure of workspace details.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The when-to-run guidance uses ambiguous conditions such as 'periodically' and requests to 'review', 'audit', or 'check' workspace without strong boundaries. This can make the skill fire opportunistically during normal conversation or maintenance flows, expanding its operational scope beyond what the user explicitly requested.

Session Persistence

Medium
Category
Rogue Agent
Content
- [ ] Old daily files reviewed and distilled to MEMORY.md periodically
- [ ] No sensitive data (API keys, passwords) in memory files

**Automatic Memory Flush:** OpenClaw triggers a silent agent turn before session compaction to write durable memories. The agent receives a prompt to flush important context to `memory/YYYY-MM-DD.md`. This is automatic — no action needed, but be aware your context WILL be compacted after ~180k tokens.

### 4. Vector Search Alignment
Confidence
78% confidence
Finding
write durable memories. The agent receives a prompt to flush important context to `memory/YYYY-MM-DD.md`. This is automatic — no action needed, but be aware your context WILL be compacted after ~180k

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.prompt_injection_instructions

Prompt-injection style instruction pattern detected.

Warn
Code
suspicious.prompt_injection_instructions
Location
references/openclaw-conventions.md:168