issue-hunter

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward GitHub issue-triage helper that uses authenticated GitHub read commands and creates a local analysis report.

Install only if you are comfortable letting the agent use your authenticated GitHub CLI for read-oriented issue and PR queries. Use a least-privileged GitHub token where possible, avoid pasting broad long-lived tokens into chat, and ask the agent to check before overwriting issue-analysis.md if that file may already matter in your workspace.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Low
Confidence
88% confidence
Finding
The skill instructs the agent to write `{workspace}/issue-analysis.md` but does not require any user notice, confirmation, or overwrite safeguards. This can cause unintended modification or replacement of an existing file in the workspace, which is especially relevant because the skill is designed to run against arbitrary repositories and workspaces where that path may already be meaningful.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal