ClawHub
Back to skill

Security audit

Github Api Toolkit Free

Security checks across malware telemetry and agentic risk

Overview

This skill is a GitHub API helper, but it gives agents broad GitHub write examples without enough built-in guardrails for high-impact actions like merging PRs or changing repositories.

Install only if you intend agents to help operate GitHub through API calls. Use the narrowest GitHub token scopes possible, avoid granting workflow unless needed, and require confirmation before any create, update, close, comment, or merge action against a real repository.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Confidence
82% confidence
Finding
The trigger conditions are broad enough to activate this skill for many generic integration requests, increasing the chance an agent selects a state-changing GitHub automation tool in contexts where the user did not clearly intend repository modifications. Because the skill includes write and exec capabilities, overbroad routing raises the risk of unintended repository, issue, or PR changes.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill documents repository-creating and repository-modifying operations without prominent warnings or confirmation requirements. In an agent context, examples of direct POST/PATCH calls can normalize destructive or irreversible changes and lead to accidental execution against real repositories.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The skill includes issue closure, commenting, PR creation, and PR merge operations without explicit caution about workflow impact or irreversibility. In an agent-enabled environment, this increases the risk of accidental closure of issues, unwanted PRs, or unintended merges that affect production workflows and collaboration history.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.