Back to skill

Security audit

wx-mp-write

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed WeChat public-account writing helper with search and optional draft upload; no hidden persistence or code execution was found.

Before using the publish flow, review the generated title, author, abstract, article body, and image URLs, and only send them to WeChat when you intend to create a draft there. Avoid putting sensitive internal drafts, customer data, or private topics into Tavily search queries unless you are comfortable sharing those terms with that service.

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
  • Rogue AgentSelf-Modification, Session Persistence
  • 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
Findings (7)

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 2)May include surrounding context.

md
---
name: wx-mp-write
description: 微信公众号文章全流程写作助手。支持热点选题、文章撰写、AI去味润色(去除机器感)、配图建议、一键发布到公众号草稿箱。适用于技术博主、自媒体创作者、企业公众号运营等场景。当用户需要写公众号文章、润色内容、寻找配图或发布到微信时使用此skill。
---

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 2)May include surrounding context.

md
---
name: wx-mp-write
description: 微信公众号文章全流程写作助手。支持热点选题、文章撰写、AI去味润色(去除机器感)、配图建议、一键发布到公众号草稿箱。适用于技术博主、自媒体创作者、企业公众号运营等场景。当用户需要写公众号文章、润色内容、寻找配图或发布到微信时使用此skill。
---

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

技能描述“当用户需要写公众号文章、润色内容、寻找配图或发布到微信时使用此skill”覆盖面很广,容易与普通写作、润色、发文相关对话重叠,导致误触发。误触发后,代理可能在用户未明确要求时进入外部搜索或发布流程,扩大数据暴露和误操作风险。

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

文档定义了“调用 wechat-publisher 发布到草稿箱”的外部写操作,但没有要求在执行前获得用户明确确认,也没有提示会把内容发送到第三方平台。对外发布即使只是草稿箱,仍可能造成敏感内容误上传、品牌风险或运营事故。

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill description is broad enough that common requests about writing, polishing, images, or publishing to WeChat could trigger it unexpectedly. Because this skill can progress from content generation into external publication workflows, accidental activation may cause unintended handling of user content or prepare content for release without the user clearly intending to use this specific skill.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill describes sending generated Markdown to an external publishing skill for placement in a WeChat public account draft box, but it does not clearly warn users that their content will be transmitted to an external integration. In a writing assistant context, users may include unpublished business content, personal anecdotes, or sensitive drafts, so failing to disclose this transfer increases the risk of inadvertent data exposure or unintended publication actions.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

技能要求配置 TAVILY_API_KEY,并引导用户在环境变量或本地配置文件中保存凭据,但没有说明密钥应如何最小化使用、避免回显,或搜索请求可能带来的隐私泄露。若用户将敏感草稿、客户信息或内部主题直接用于搜索,可能把数据发送给外部服务。

Content

No source excerpt is available for this finding.