Security audit
Security checks for vulnerabilities and agentic risk
This skill is a local reflection tool that reads prior conversation content and stores limited notes under ~/meditate, with no evidence of network use, hidden execution, or destructive behavior.
Before installing, understand that this skill is designed to analyze prior conversations and keep derived reflection notes locally in ~/meditate. Review or delete that directory if you do not want persistent profile, topic, feedback, or archived insight data retained.
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
# User Profile
## Detected Type
<!-- entrepreneur | developer | creative | personal | system | unknown -->
Type: unknown
Confidence: low
Last updated: never
The file-level documentation states or implies that reflections are sandboxed, yet all templates are copied into persistent files in the user's home directory. This is dangerous because users may rely on the stronger privacy claim and provide personal feedback or topics they would not share if they knew the data was stored long-term outside a sandbox.
The skill description promises 'sandboxed reflections during idle time', but the template explicitly instructs persistent storage of user profile, topics, insights, and feedback under ~/meditate. This creates a mismatch between documented privacy expectations and actual data handling, which can expose sensitive behavioral or preference data to other local processes, backups, or users with access to the home directory.
The skill explicitly documents persistent storage under ~/meditate/ for conversation-derived profile, topics, insights, feedback, and archives, but it does not present a clear user-facing warning or consent mechanism for this retention. Even though the storage is local and scoped, users may not expect their conversations to be analyzed and written to disk, creating a privacy and transparency issue.
The scope states that the skill reads conversation history and stores insights in a local queue, but it does not clearly warn that this entails ongoing analysis of user content and retention of derived data over time. This is a real privacy weakness because users may interpret the skill as passive reflection rather than persistent profiling and storage.
No suspicious patterns detected.