ClawHub
Back to skill

Security audit

Quit Sponsor

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed quit-smoking support skill that uses local memory/logging for continuity, with sensitive-data risks users should understand before enabling it.

Install only if you want an AI agent to keep a private, ongoing quit-smoking log. Use local/private memory, review what is being stored, and request deletion if you stop using it. Do not rely on it as medical or crisis care; use clinicians, emergency services, or crisis lines for red flags or acute distress.

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 (4)

Vague Triggers

Medium
Confidence
88% confidence
Finding
The activation description is extremely broad and can trigger on many ordinary conversations about smoking, relapse, silence, or habit change. Because the skill then adopts a sponsor role with memory, logging, and quasi-therapeutic protocols, accidental activation could cause unconsented behavioral coaching and sensitive data handling in contexts where the user did not explicitly ask for that level of intervention.

Ssd 3

Medium
Confidence
95% confidence
Finding
The skill explicitly assumes persistent memory and directs maintaining a logbook across sessions, including contracts, risk maps, slips, and quit history. This creates broad retention of sensitive health and behavioral data, increasing the blast radius of any memory leakage, unauthorized access, model misuse, or unintended reuse in later conversations.

Ssd 3

High
Confidence
98% confidence
Finding
This section instructs the agent to store sensitive health and personal history in memory across sessions, including mental-health history, medications, living situation, and physical health flags. Even though it says to store flags rather than narratives, these are still highly sensitive health inferences whose persistence can expose users to privacy harm, profiling, or inappropriate downstream use.

Ssd 3

High
Confidence
97% confidence
Finding
The protocol directs detailed longitudinal tracking of slips, reactions, confession timing, cumulative counters, and confirmation status, creating a rich behavioral dossier over time. In the context of addiction support, these records can reveal health status, routines, emotional states, social relationships, and relapse patterns, making them highly sensitive and particularly damaging if exposed or repurposed.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.