Research to Sheets

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed workflow for researching web data and writing confirmed results to Google Sheets through ClawLink.

Before installing, confirm you trust the ClawLink plugin and are comfortable enabling it in OpenClaw. Use the skill’s preview and confirmation steps before any spreadsheet write, especially for existing or shared sheets.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (2)

Context-Inappropriate Capability

Medium
Confidence
94% confidence
Finding
The skill directs the agent to install a plugin, widen the tool allowlist, and restart the gateway, which are privileged local administrative actions unrelated to the core task of researching websites and writing rows to Google Sheets. This expands the agent's capabilities and changes the host environment, creating a supply-chain and privilege-expansion risk if the plugin or configuration is unsafe or compromised.

Context-Inappropriate Capability

Medium
Confidence
92% confidence
Finding
The troubleshooting section includes additional host-level commands to inspect plugins, change configuration, and restart the gateway. Even as fallback guidance, these instructions normalize privileged environment changes beyond the skill's stated purpose and could be abused to persistently alter the agent runtime or enable unintended tools.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal