ClawHub
Back to skill

Security audit

Maya Loop

Security checks across malware telemetry and agentic risk

Overview

This skill is a development workflow, but it asks agents to activate broadly and run autonomous operational loops with limited user control.

Install only if you intentionally want an opinionated engineering workflow that may guide agents during code changes, deployments, backups, server checks, and persistent logging. Review and narrow the trigger conditions, automation steps, project paths, backup destinations, and any scheduled loops before enabling it in sensitive repositories or production environments.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (3)

Vague Triggers

High
Confidence
98% confidence
Finding
The trigger conditions are extremely broad and include ordinary development phrases, any file modification, and any product development scenario. This can cause the skill to activate without explicit user consent and steer the agent into planning or operational behavior in contexts where the user only intended discussion, increasing the risk of unintended system-affecting actions.

Vague Triggers

High
Confidence
99% confidence
Finding
The main instructions mandate automatic activation whenever common development actions or keywords appear, and explicitly say not to wait or ask. In a skill that governs code changes, deployment, and rollback-sensitive operations, this ambiguity and overbreadth can override normal consent boundaries and lead to unauthorized or premature execution flows.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The front matter advertises broad automatic triggering for write, edit, bash, deployment, and autonomous operations, but does not provide a clear user-facing warning that the skill may initiate system-affecting behaviors. This reduces informed consent and increases the chance that users enable or rely on the skill without understanding its operational reach.

VirusTotal

62/62 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.