ClawHub
Back to skill

Security audit

Data Analysis Workflow

Security checks for vulnerabilities and agentic risk

Overview

This is a low-risk data analysis workflow scaffold that reads user-specified datasets and prints basic inspection output, with some overstated documentation.

Before installing, understand that this appears to be a workflow guide plus a basic dataset inspection script, not a full automated statistical analysis and reporting tool. Use it only with files you intend to analyze, and expect Chinese-language output unless the skill is updated.

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
  • 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 (5)

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding
The description claims a complete standardized data analysis workflow integrating data-analysis, statistical-analysis, scientific-visualization, and other skills across six stages from import to reporting. However, the supplied code mainly defines stage metadata, loads tabular data, prints shape/types/missing values/describe(), and outputs guidance on what external skills should be used next. There is no implementation of cleaning/preprocessing, inferential statistical testing, visualization generation, or report generation. This is not a harmful undeclared capability issue; rather, the primary mismatch is overstatement of what the code actually accomplishes. The code represents a workflow scaffold or launcher/orchestrator for stage 1 plus recommendations, not a complete end-to-end analysis workflow as described.

Natural-Language Policy Violations

Medium
Confidence
97% confidence
Finding
The file presents the operational skill documentation almost entirely in Chinese, despite the manifest metadata and section headings beginning in English. There is no statement that the skill is region-specific or that users may choose another language, which creates a natural-language policy concern under the language/locale rule.

Natural-Language Policy Violations

Medium
Confidence
88% confidence
Finding
The skill description is written entirely in Chinese and presents the workflow as standardized in that language, while the manifest also includes English trigger keywords. This creates a language/locale inconsistency and may effectively force a specific language experience without explicit user opt-in or documented justification.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The trigger list includes generic phrases like "data analysis" and "statistical analysis," which are common in ordinary user requests and can cause this workflow to activate too broadly. Over-broad invocation increases the chance of unintended routing, unexpected tool/skill chaining, and reduced user control over which analysis workflow is applied.

Natural-Language Policy Violations

Medium
Confidence
92% confidence
Finding
The module description explicitly labels the workflow as standardized in Chinese, and the user-facing strings throughout the script are presented only in Chinese. This creates a language/locale constraint without any documented opt-in or alternative, which fits the policy category for forced language behavior.

Static analysis

No suspicious patterns detected.