Vector Db Manager Free

• Outcome: Review
• Latest audit: Jul 23, 2026 · 4:01 PM UTC
• Version: 1.0.0

Overview

The skill is a coherent vector database management guide, but it gives agents broad database-changing authority without enough scoping or confirmation safeguards.

Review this skill before installing in environments with real databases. Use it only with explicit target database details, prefer read-only or test credentials first, and require manual confirmation before delete, drop, rebuild, overwrite, import, or export actions.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Behavioral AST: exec() Call, eval() Call, Dynamic Import
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Vague Triggers

Medium

• Confidence: 87% confidence
• Finding: The activation scope is broad enough to match generic database and SQL tasks, which can cause the skill to be invoked outside its intended vector-database context. Because the skill exposes exec and database-management behavior, unintended activation could lead an agent to perform risky operations against the wrong system or with insufficient user confirmation.

Vague Triggers

Medium

• Confidence: 90% confidence
• Finding: The keyword coverage list is excessively broad and mechanically derived, increasing the chance that an agent routes unrelated requests to this skill. In an exec-enabled skill that can manage indexes, collections, imports, and deletions, accidental invocation expands the attack surface and raises the risk of unintended data modification.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The skill advertises destructive actions like deleting indexes, collections, and partitions without prominent warnings about irreversible data loss or a requirement for confirmation. In the context of database administration with exec access, this omission is dangerous because an agent may carry out destructive operations on production or cloud-backed data without the user appreciating the risk.

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: Import/export/save/convert operations are described as supported outputs without warning about overwriting files, exfiltrating data, or exposing sensitive embeddings and metadata. Since embeddings and associated metadata can contain proprietary or personal information, silent export or save behavior can create confidentiality and integrity risks.

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: This skill is exec-enabled and intended to manage local and cloud vector databases, yet it does not prominently warn that commands may run against real external services using configured credentials. In this context, omission of environment and side-effect warnings is dangerous because users may unknowingly authorize changes to live databases or trigger networked operations against cloud resources.

VirusTotal

VirusTotal findings are pending for this skill version.

Static analysis

No suspicious patterns detected.