Back to skill

Security audit

Shows

Security checks for vulnerabilities and agentic risk

Overview

This media-tracking skill is mostly coherent, but it can persist family and child-specific viewing notes with broad triggers and limited privacy controls.

Install only if you are comfortable with a local ~/shows folder retaining viewing history, preferences, family-member details, and possibly child-specific reaction notes. Prefer aliases or age ranges for children, avoid recording unnecessary sensitive reactions, and review or delete the local files periodically.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
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 (5)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill activates on very broad, everyday phrases like mentions of watching something or asking what to watch, without clear boundaries that confirm the user intends to use this specific tracking skill. That can cause unintended collection or modification of personal viewing history, ratings, and household-related data from ordinary conversation rather than explicit opt-in use.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill instructs storing persistent viewing records, ratings, notes, and family-member information in local files without disclosing that this data will be retained. Even if the data is not highly sensitive in isolation, it can reveal behavioral patterns, household composition, preferences, and routines, creating a privacy risk if stored unexpectedly or accessed by others on the same system.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The skill uses broad trigger phrases like "what should I watch?" and "recommend something," which overlap with common user requests and can cause the skill to activate in situations where the user did not explicitly intend to invoke it. This can lead to overbroad interception of conversational requests, unintended behavior precedence over other skills, and possible misuse of personal preference/history data in the wrong context.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill explicitly models storage of children's names, ages, and individualized viewing history, which is personal data and can become sensitive when tied to minors. Even though the file does not exfiltrate data, normalizing collection and persistence of child-related data without privacy guidance, minimization, retention limits, or consent warnings creates avoidable privacy risk if the notes are shared, synced, or accessed by others.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill encourages logging behavioral and emotional reaction notes about children, such as fear responses and nightmares, which can constitute sensitive profiling data. In the context of a family-tracking skill, this makes the issue more concerning because the system is not merely recommending content but persisting child-specific psychological observations without caution about confidentiality, necessity, or parental review.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.