Security audit

voice-learn

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed local voice-learning helper, but users should understand that it stores writing snapshots in their local Claude profile.

Install this only if you are comfortable with the skill keeping local copies of generated drafts, reviewed drafts, and your final edits under the Claude voice-profile directory. Avoid using it on confidential writing unless you have checked where that profile directory is stored or synced, and review proposed profile changes before approving them.

SkillSpector

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 (3)

Vague Triggers

Medium

Confidence: 92% confidence
Finding: The trigger list includes broad everyday terms like "voice," "learning," "improvement," "feedback," and "iteration," which can cause the skill to activate in many unrelated conversations. Because this skill reads and updates profile and learning files, accidental invocation could expose user content or cause unintended modifications to long-term voice-profile state.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The module explicitly captures and stores pre-review, post-review, and post-edit content, which can include sensitive user-generated text and manual edits, but it provides no privacy notice, consent step, access control guidance, or retention safeguards at the point of collection. In a voice-learning workflow, these snapshots may preserve drafts, rejected content, and edits that reveal confidential information or writing habits, increasing privacy and data exposure risk if the local profile directory is accessed, synced, or backed up.

Missing User Warnings

Low

Confidence: 86% confidence
Finding: The cleanup guidance allows snapshots to be archived or deleted but does not warn users that archived data may still persist elsewhere or that deletion may be irreversible. Because the module handles accumulated writing snapshots, unclear retention and disposal behavior can lead to mistaken assumptions about privacy, recoverability, or compliance with user expectations.

Static analysis

No suspicious patterns detected.