Back to skill

Security audit

自然中文协议 (Natural Chinese Protocol)

Security checks for vulnerabilities and agentic risk

Overview

This is a Chinese writing-style skill that rewrites text to sound more natural, with no code execution, persistence, credential access, or hidden data handling.

Install this only if you want an optional Chinese text naturalization/editing profile. Avoid using it to hide required AI-use disclosures, alter legal/compliance statements, or rewrite text when the user asked for a different tone or language.

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
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (4)

Natural-Language Policy Violations

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill is framed as a core protocol to 'eliminate AI traces' in Chinese text generation, which pushes a mandatory stylistic transformation without explicit user opt-in. In an agent setting, this can cause unauthorized rewriting, misrepresentation of authorship/origin, and conflict with user-requested tone, compliance, or disclosure requirements.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

High
Category
Not specified by scanner
Confidence
93% confidence
Finding

The line uses mandatory language ('直接挂载声明') instructing operators to apply the protocol in prompts/workflows, effectively enforcing a specific Chinese writing style without per-request opt-in. In context, the protocol is designed to conceal machine-like characteristics, making the style imposition more dangerous because it can systematically alter outputs in ways users did not request and may undermine transparency.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The README says the protocol is "widely applicable to all general Chinese writing and communication scenarios" and can be invoked by directly adding a prompt to apply it. This lacks clear scope boundaries or exclusion conditions, making the trigger and activation context broad enough to overlap with many ordinary writing tasks.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The instruction states the agent is responsible for editing in general Chinese writing scenarios, which imposes a specific language/locale behavior. Under the policy, forcing a language without user opt-in is a natural-language policy violation unless the constraint is explicitly justified as region-specific or optional.

Content

No source excerpt is available for this finding.