ClawHub
Back to skill

Security audit

精简版Sqlite

Security checks across malware telemetry and agentic risk

Overview

This SQLite helper appears non-malicious, but it asks for broad database-changing capability while leaving API key use, data flow, and destructive-operation safeguards unclear.

Review this skill before installing in workflows that touch important or sensitive databases. Use it only on intended SQLite files, avoid giving it broad workspace access, confirm any update or delete operation explicitly, keep backups, and clarify whether any LLM/API integration can see database contents.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Intent-Code Divergence

Medium
Confidence
95% confidence
Finding
The skill markets itself as a lightweight local SQLite capability, but elsewhere requires an API key and an LLM API, creating a misleading trust boundary. Users may provide sensitive local data under the assumption it stays local when the documented dependencies imply potential remote processing or telemetry.

Intent-Code Divergence

Low
Confidence
90% confidence
Finding
The security/compliance table claims protections such as encryption in storage and transit, but the file provides no implementation details or configuration proving those controls exist. Unsupported security claims can cause operators to overtrust the skill and store sensitive data without applying needed compensating controls.

Vague Triggers

Medium
Confidence
84% confidence
Finding
The activation guidance is broad and repetitive, effectively suggesting use for generic database operations without clear boundaries, exclusions, or safety conditions. In an agent environment, vague triggers increase the chance the skill is invoked on inappropriate tasks or sensitive datasets without deliberate user intent.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The usage example includes a direct delete operation without any warning, confirmation flow, backup guidance, or mention that the action is irreversible. In an agent-assisted workflow, this can normalize destructive behavior and lead to accidental data loss if copied or auto-generated into real tasks.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.