Back to skill

Security audit

销售干货日更流水线

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Chinese sales/fraud-prevention article generator that writes generated articles and a local HTML library, with notable but manageable concerns around broad triggers and off-platform paid-upgrade instructions.

Install only if you want a Chinese-language daily article workflow that saves Markdown files and updates a local HTML article library. Use a dedicated article folder, review generated content before publishing, and treat the WeChat paid-upgrade/payment instructions as external to platform protections.

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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The summary/description says the free version depends on manual duplicate checking and manual pre-publication self-checks, with automation only in the paid version. However, the document later instructs the agent to run an automation script for library insertion and includes embedded sales/contact instructions for purchasing the paid toolkit, which goes beyond the claimed free/manual-only framing and creates intent divergence within the documentation itself.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger phrases are broad enough that ordinary user language such as 'write today's sales article' or 'daily write one' could invoke the skill unintentionally. Because the skill can proceed to generate files, remember directories, and update a local article library, accidental activation may cause unreviewed content creation or persistence without clear user intent.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill performs persistent local side effects: it saves generated content to disk, increments filenames, and invokes a script to update a local full-text library, yet the description does not clearly warn users about these automatic writes. In agent environments, undisclosed file modification is risky because users may believe they are only requesting content generation, not authorizing filesystem changes.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

The skill embeds a direct WeChat contact and payment workflow, encouraging users to move to an external channel and share payment proof, without any privacy or safety warning. While not inherently malicious, this creates social-engineering and privacy risks by normalizing off-platform contact and disclosure of personal/payment information inside a skill document.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
94% confidence
Finding

The generated default HTML template sets lang="zh" and all user-facing strings are fixed in Chinese. This is a natural-language/locale policy concern because the skill forces a specific language/locale without any opt-in, configuration, or explanation that the tool is intentionally region-specific.

Content

No source excerpt is available for this finding.