ClawHub

CI Whisperer

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed GitHub Actions troubleshooting helper with read-only defaults and an explicitly gated PR-creation mode.

Install only if you are comfortable with the agent using your local GitHub CLI identity to inspect workflow runs and logs. Keep CI_WHISPERER_WRITE unset unless you deliberately want it to create a branch or PR, and review log excerpts because CI output can contain sensitive data.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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 (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
87% confidence
Finding
The skill invokes shell-based tooling (`gh`) and references bundled scripts, but the manifest does not declare corresponding permissions or clearly constrain execution scope. This creates a trust and review gap: consumers may authorize or run the skill without understanding that it can execute local commands and access repository/workflow data through the authenticated GitHub CLI.

Tp4

High
Category
MCP Tool Poisoning
Confidence
82% confidence
Finding
A description-behavior mismatch is a security-relevant integrity problem because users and reviewers rely on the manifest to understand what the skill will do. Here, the skill claims safe, concise CI failure analysis, but the documented workflow also allows broader repository/run enumeration and references optional write actions, while the actual described behavior does not guarantee root-cause analysis or remediation quality; this can lead to over-trust and unsafe invocation in sensitive repos.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal