Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md 每个 Skill 是一个包含 `SKILL.md` 的目录。通过 YAML frontmatter 声明元数据。
Security audit
openclaw-overview
Security checks for vulnerabilities and agentic risk
Overview
This is an informational OpenClaw overview skill, not an executable skill that installs persistence or takes actions by itself.
Before installing, understand that this skill teaches an agent about OpenClaw's powerful runtime features, including shell commands, file edits, messaging, device nodes, cron, memory, and gateway configuration. It appears safe as documentation, but users should still review any later agent-proposed command, device pairing, cron job, or configuration change before allowing it.
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
- Rogue AgentSelf-Modification, Session Persistence
- 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 (6)
Missing User Warnings
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The markdown states that paired node devices can provide camera access, screen recording, location, voice wake, and notification reading. These are privacy-sensitive capabilities, but the skill description does not include a user-facing warning about the potential collection or exposure of personal/device data.
- Content
Session Persistence
Medium
- Category
- Rogue Agent
- Confidence
- 60% confidence
- Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.
- Content
openclaw update # 检查并更新
备份
openclaw backup create # 创建备份
text ---
Session Persistence
Medium
- Category
- Rogue Agent
- Confidence
- 60% confidence
- Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.
- Content
openclaw update # 检查并更新
备份
openclaw backup create # 创建备份
text ---
Missing User Warnings
Low
- Category
- Not specified by scanner
- Confidence
- 87% confidence
- Finding
The tool overview advertises
exec/processfor shell command execution and process management. While this is part of the platform capability overview, the markdown does not warn that these operations can modify the system, run destructive commands, or affect local integrity.- Content
Missing User Warnings
Low
- Category
- Not specified by scanner
- Confidence
- 84% confidence
- Finding
The markdown explicitly lists
writeandeditcapabilities for file modification, and later references configuration read/write actions. The description does not provide any warning that these actions can alter or overwrite user files and configuration.- Content
Static analysis
No suspicious patterns detected.