ClawHub
Back to skill

Security audit

Rag Construction

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent construction-document RAG helper, with expected file/data handling but no evidence of hidden execution, exfiltration, persistence, or destructive behavior.

Install only for construction document workflows and avoid indexing confidential project files unless you intend the agent to use and possibly export their full content. Keep access limited to the relevant project folder and review exports before sharing them.

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
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill implements a knowledge base that stores full document content in memory and exposes it via export_knowledge_base(), including metadata and sources, but the documentation does not warn about retention, sensitivity, or data handling risks. In a construction context, documents may contain contracts, RFIs, safety reports, or proprietary project data, so users could unintentionally ingest and later expose sensitive information.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The instructions expand the skill from a narrowly scoped construction RAG assistant into a more general document-generation and data-processing workflow. This increases the chance the agent will act on arbitrary file paths, transform user-supplied data, or perform broader tasks than intended, weakening scope boundaries and making misuse or unsafe tool invocation more likely.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

SQP-3 applies to all file types and covers language or locale policy issues. The line presents the title in Russian and English as a fixed reference, with no indication that language selection is optional or contextually required.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.