Security audit

AI Chatbot Service

Security checks for vulnerabilities and agentic risk

Overview

This skill is a simple local Chinese-language FAQ chatbot with documented admin commands and no evidence of network access, persistence, credential use, or hidden execution.

Install only if a Chinese-language local FAQ/customer-service helper fits your use case. Avoid entering unnecessary personal, order, payment, or contact details because recent chat snippets may be shown by the stats command within the same runtime, and use clear when finished in shared environments.

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

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (6)

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 87% confidence
Finding: The skill documents adding FAQ entries and maintaining conversation history, but it does not warn that these may contain personal, transactional, or contact data that could later be exposed to operators, logs, or other users if mishandled. In a customer-service context, users commonly share addresses, phone numbers, order details, and complaint content, so omission of privacy and retention guidance increases the chance of accidental sensitive-data disclosure.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: This file presents the tool description and all user-facing interactions in Chinese, indicating a fixed language behavior. Under the policy rule, forcing a specific language without user opt-in is a natural-language policy violation unless the locale restriction is explicitly justified, which it is not here.
Content: No source excerpt is available for this finding.

Ssd 3

Medium

Category: Not specified by scanner
Confidence: 90% confidence
Finding: The skill stores raw user messages in a global conversation history and later re-displays them through the stats command. In a shared runtime, later users or operators could see previous users' sensitive inputs, creating an information disclosure risk; the undocumented inspection/statistics context makes this more dangerous because customer service conversations often contain order, payment, or personal details.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: The manifest frames the skill as an intelligent customer-service system focused on answering questions and dialogue handling. However, the add command lets a caller modify the FAQ knowledge base at runtime, which is a separate content-management capability rather than an obvious implementation detail of answering customer questions.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 87% confidence
Finding: The description, examples, intent keywords, and usage guidance are written entirely in Chinese, and the skill does not state that other languages are supported or let users opt into a preferred language. Under the language/locale policy, forcing a specific language without user opt-in can be a policy violation unless clearly justified as region-specific.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Low

Category: Not specified by scanner
Confidence: 84% confidence
Finding: The manifest mentions FAQ auto-reply, intent recognition, and multi-turn conversation, but the code also provides list, stats, and clear commands for enumerating the FAQ database, viewing conversation metrics, and clearing history. These are operational/admin functions beyond the user-facing service behavior described in the manifest.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.