Security audit

Data Analyst Partner

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent analytics helper skill for Grafana and ClickHouse workflows, with no hidden execution, persistence, or unrelated access behavior found.

Install this skill only in environments where the agent is already allowed to read the relevant Grafana dashboards or ClickHouse-backed business metrics. Teams should confirm the preferred response language and keep direct SQL access governed by existing data permissions.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (3)

Vague Triggers

Medium

Confidence: 96% confidence
Finding: The skill description is broad enough to activate on many ordinary analytics-related requests, which can cause the agent to invoke this skill in situations where a narrower or safer skill would be more appropriate. Overbroad routing is risky because it can expand the skill’s operational scope, increase unintended access to Grafana/ClickHouse-backed analysis flows, and make least-privilege enforcement harder.

Natural-Language Policy Violations

Medium

Confidence: 86% confidence
Finding: The answer template strongly biases output toward Chinese without stating that the assistant should follow the user’s language preference. This can create usability and operational risk by causing responses in an unexpected language, which may lead to misunderstanding of analytical conclusions or accidental disclosure to the wrong audience in multilingual environments.

Natural-Language Policy Violations

Low

Confidence: 89% confidence
Finding: This markdown file includes an example report skeleton that uses Chinese section labels (总览, 变化, 异常, 建议) while the rest of the template is written in English. That creates an implicit language expectation without stating that the skill is region-specific or giving users a language/locale choice.

Static analysis

No suspicious patterns detected.