ClawHub
Back to skill

Security audit

卖家之家(跨境电商)技能商城(skillhub/clawhub/搜技能/热门技能/免费技能/付费技能)

Security checks for vulnerabilities and agentic risk

Overview

The skill is a disclosed marketplace helper for searching and submitting MJZJ skills, with expected API-key use for uploads and review submission.

Install this only if you intend to use MJZJ SkillHub and are comfortable providing an MJZJ API key. Review the skill metadata, source URL, labels, price type, tags, and any cover image before asking the agent to submit a publishing application, since that sends the approved content to MJZJ for review.

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
  • 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 (8)

Natural-Language Policy Violations

Medium
Confidence
96% confidence
Finding
The file’s user-facing description is written only in Chinese, which imposes a specific language on users without documenting any opt-in or alternative locale. This matches the policy category for language or locale constraints that are not optional or justified.

Natural-Language Policy Violations

Medium
Confidence
91% confidence
Finding
The entire SKILL.md, including trigger guidance and usage instructions, is presented exclusively in Chinese and does not indicate that users may choose another language. For a general-purpose skill, this is a natural-language locale policy issue unless the region/language limitation is clearly justified.

External Transmission

Medium
Category
Data Exfiltration
Content
### 1) 获取技能标签分组（公开）

```bash
curl -X GET "https://data.mjzj.com/api/skill/groupLabels" \
  -H "Content-Type: application/json"
```
Confidence
60% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
### 1) 获取技能标签分组（公开）

```bash
curl -X GET "https://data.mjzj.com/api/skill/groupLabels" \
  -H "Content-Type: application/json"
```
Confidence
50% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
### 1) 获取技能标签分组（公开）

```bash
curl -X GET "https://data.mjzj.com/api/skill/groupLabels" \
  -H "Content-Type: application/json"
```
Confidence
50% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
### 1) 获取技能标签分组（公开）

```bash
curl -X GET "https://data.mjzj.com/api/skill/groupLabels" \
  -H "Content-Type: application/json"
```
Confidence
50% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
自定义封面时，先申请上传临时文件：

```bash
curl -X POST "https://data.mjzj.com/api/common/applyUploadTempFile" \
  -H "Authorization: Bearer $MJZJ_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
Confidence
82% confidence
Finding
This workflow requests a temporary upload URL and then instructs the client to PUT a file to that returned putUrl. Because the skill does not state any validation or allowlisting of the returned upload host, a compromised or malicious upstream service could supply an arbitrary destination and cause unintended transmission of user-provided file content and authorization-scoped workflow data to an attacker-controlled endpoint.

External Transmission

Medium
Category
Data Exfiltration
Content
然后把返回的 `path` 作为 `coverFile` 提交：

```bash
curl -X POST "https://data.mjzj.com/api/skillManage/applyNew" \
  -H "Authorization: Bearer $MJZJ_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
Confidence
73% confidence
Finding
This authenticated POST submits skill publication data, including metadata and a server-issued temporary file path, to an external service using the MJZJ API key. While such transmission is part of the feature, it handles authenticated content creation on a third-party platform; if invoked without clear user intent or with insufficient confirmation, it could enable unauthorized publication or disclosure of user-supplied content to the remote service.

Static analysis

No suspicious patterns detected.