ClawHub
Back to skill

Security audit

Session Memory Manual

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed manual memory-search skill, but it can read prior OpenClaw sessions and write user-approved notes into shared memory.

Install this only if you want manual cross-session lookup. When using it, specify exact keywords, sessions, or time ranges; review results before approving any memory write; and avoid storing passwords, API keys, private chats, or sensitive business details in shared memory.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Vague Triggers

Medium
Confidence
88% confidence
Finding
The skill claims to be manual-only, but its trigger guidance still uses broad natural-language examples like '帮我查' and '搜一下', which can be interpreted loosely and lead an agent to perform cross-session retrieval on ambiguous user requests. Because the skill operates on shared memory and session history, overly permissive triggering increases the chance of unintended access to prior conversations or private data despite the stated privacy safeguards.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.