Hologres Query Optimizer

• Outcome: Pass
• Latest audit: May 28, 2026, 4:36 AM
• Version:
  0.2.0

Overview

This is a documentation-only Hologres query-tuning skill whose database commands are relevant to its purpose, but users should treat the examples as real operational changes.

Install only if you intend to use Hologres query tuning guidance. Verify hologres-cli before installing it, use least-privilege database credentials, and require explicit review before running EXPLAIN ANALYZE, ANALYZE, set_table_property, or persistent hologres guc set commands, especially on production databases.

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
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Description-Behavior Mismatch

Medium

• Confidence: 92% confidence
• Finding: The skill is presented as an analyzer/optimizer, but its instructions go beyond passive analysis and direct users to execute SQL and inspect or modify runtime/database settings via CLI. That scope expansion is risky because an agent or user invoking an 'analysis' skill may not expect actions with operational side effects, especially when EXPLAIN ANALYZE executes queries and GUC operations can change behavior.

Description-Behavior Mismatch

Medium

• Confidence: 95% confidence
• Finding: The documentation includes persistent database-level configuration changes using `hologres guc set`, which materially exceeds a read-only optimization/advisory role. Persistent tuning changes can alter query behavior for other workloads, cause outages or regressions, and create a privilege boundary issue if an agent applies them automatically.

Vague Triggers

Medium

• Confidence: 84% confidence
• Finding: The trigger phrases are broad, generic SQL-performance terms that are likely to match many ordinary database support requests. Because this skill can lead to query execution and configuration recommendations, overly broad invocation increases the chance that it activates in the wrong context and prompts risky actions without the user intending to use this specific skill.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The skill describes using `EXPLAIN ANALYZE`, SQL execution, and GUC parameter operations without a clear, prominent warning that these actions can execute user queries, consume resources, and potentially change database behavior. In a production environment, missing side-effect warnings can lead users or agents to run expensive statements or alter settings under the mistaken assumption that the skill is purely observational.

Missing User Warnings

Low

• Confidence: 91% confidence
• Finding: The document provides persistent database-level `hologres guc set` examples without clearly warning that these changes outlive the current session and can affect all users and workloads on the database. In a query-optimization skill, operators may copy these commands directly, causing unintended performance regressions, memory pressure, or broader service impact if experimental or aggressive settings are left enabled.

VirusTotal

VirusTotal findings are pending for this skill version.