Back to skill

Security audit

Bilingual Mail Translator

Security checks for vulnerabilities and agentic risk

Overview

This markdown-only skill coherently documents a Chinese bilingual email translation workflow, with no hidden code, persistence, or install-time behavior found.

Install only if you want agents to help design or maintain a Chinese bilingual email translation workflow. Because the workflow may process raw email content, use approved LLM runtimes, avoid unnecessary retention or logging, and redact sensitive mail when possible.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill is described with broad trigger language around building or improving email translation and notification workflows, which can cause over-activation on general email tasks. In an agent setting, that increases the chance that raw email content is routed through this skill unnecessarily, expanding exposure of sensitive message data and causing the skill to influence tasks beyond its intended scope.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill is framed as producing Chinese-friendly bilingual output by default, without stating that this should be user-selected or locale-driven. In practice, this can cause unintended transformation or disclosure of email content to additional language-processing paths, and may produce incorrect or inappropriate handling for users, regions, or messages where translation is not desired.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill explicitly instructs program flow to extract raw mail and send it to a lightweight worker or LLM runtime, but it does not warn that emails may contain sensitive personal, business, or credential-bearing information. That omission can lead operators or downstream systems to process confidential email content through external or less-trusted model infrastructure without informed consent or data-handling safeguards.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
85% confidence
Finding

The document instructs use of the label 历史邮件摘要:, which imposes a specific locale/language in the output. Because the file does not indicate user opt-in or a region-specific requirement, this appears to conflict with the policy against forcing a specific language without choice.

Content

No source excerpt is available for this finding.