Security audit

Brand Voice Profile

Security checks for vulnerabilities and agentic risk

Overview

This is a simple brand-voice writing skill whose local profile storage and style analysis are aligned with its stated purpose, with some privacy considerations users should understand.

Before installing, understand that this skill is meant to save a reusable brand voice profile locally. Do not provide private drafts, client content, or unpublished material for analysis unless you want those style details reflected in future outputs; review or delete brand-voice/profile.json if you no longer want it retained.

A.I.G

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

SkillSpector

Vulnerability Patterns

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
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (3)

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The skill directs the agent to create and update `brand-voice/profile.json` without explicitly warning the user that this causes persistent local file writes. This can lead to unintended retention of sensitive brand, audience, or writing-preference data and surprises users who expected a transient conversation rather than storage.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The skill instructs the agent to read a user's last 10 posts or articles to infer style, but it does not warn that those materials may contain sensitive or private information. If used on local drafts, unpublished articles, or mixed-content directories, the agent could ingest more data than the user intended and persist or reuse it in later outputs.

Vague Triggers

Low

Confidence: 90% confidence
Finding: The description says to use the skill for broad situations like generating content that matches your voice, onboarding a new content workflow, or ensuring consistency across platforms. These are general writing scenarios and the file does not define specific trigger phrases, scope limits, or negative examples, which could cause unintended invocation in ordinary content tasks.

Static analysis

No suspicious patterns detected.