ClawHub
Back to skill

Security audit

输入关键词，自动扫描全网用户痛点，生成PRD并输出OpenClaw标准任务JSON。

Security checks for vulnerabilities and agentic risk

Overview

This skill searches public web sources for a user-provided keyword and generates a PRD-style JSON, with its main network and API-key behavior disclosed.

Install only if you are comfortable sending search keywords to Brave Search and Hacker News Algolia. Do not use confidential product names, customer data, secrets, or internal project details as keywords, and treat the Baidu setup notes as inconsistent documentation rather than working Baidu support.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (9)

Tp4

High
Category
MCP Tool Poisoning
Confidence
90% confidence
Finding

The primary purpose broadly matches the description: it searches for pain points and outputs PRD-like JSON. However, the declaration is incomplete relative to actual behavior. The code clearly performs external network requests to Brave and Hacker News, and transmits the user keyword to those services, while declared permissions are empty. It also consumes an API key from the environment. Those are material capabilities/resources that should be declared. The extra generation of user stories, MVP features, and OpenClaw tasks is adjacent to PRD generation rather than unrelated, so it is a minor expansion, but the undeclared network/resource usage is sufficient to flag a mismatch.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · FIXES.md (reported line 179)May include surrounding context.

md
3. 配置 API Key (推荐)
   export BRAVE_API_KEY="your_api_key"
   
   获取方式: https://api.search.brave.com/

4. 测试新版本
   python3 ~/.openclaw/skills/auto-prd-hunter/scripts/search.py \

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · FIXES.md (reported line 196)May include surrounding context.

md
3. 配置 API Key (推荐)
   export BRAVE_API_KEY="your_api_key"
   
   获取方式: https://api.search.brave.com/

4. 测试新版本
   python3 ~/.openclaw/skills/auto-prd-hunter/scripts/search.py \

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 251)May include surrounding context.

md
3. 配置 API Key (推荐)
   export BRAVE_API_KEY="your_api_key"
   
   获取方式: https://api.search.brave.com/

4. 测试新版本
   python3 ~/.openclaw/skills/auto-prd-hunter/scripts/search.py \

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding

The skill declares executable capabilities involving environment variables and outbound network access, but does not define an explicit tool scope such as permissions or allowed-tools. This weakens reviewability and enforcement, making it easier for a user-invocable skill to access sensitive runtime capabilities without a clear least-privilege contract.

Content

No source excerpt is available for this finding.

Scope Creep

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The documentation instructs operators to configure and use a Baidu API key, but the manifest does not declare a Baidu network destination. This creates a discrepancy between reviewed network scope and actual intended behavior, enabling undeclared data transmission to an external service if the script supports it.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
91% confidence
Finding

The script sends the user-supplied keyword to third-party services (Brave Search and Hacker News/Algolia), which is an external data transmission. If users include confidential project names, internal problem statements, customer data, or other sensitive terms in the keyword, that information is disclosed to external providers and may be logged or retained outside the user's control. The skill context increases relevance because the tool is explicitly designed to take freeform user queries and transmit them over the network.

Content

Scanner excerpt · scripts/search.py (reported line 50)May include surrounding context.

python
results = []
    try:
        url = f"https://api.search.brave.com/res/v1/web/search?q={urllib.parse.quote(keyword)}&count={max_results}"
        req = urllib.request.Request(url)
        req.add_header("Accept", "application/json")
        req.add_header("X-Subscription-Token", api_key)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This code produces multiple user-facing fields such as user stories, feature descriptions, task titles, and summary text in Chinese only. The file does not indicate that the skill is region-specific or allow users to opt into a preferred language, which creates a language/locale policy issue.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

The cron integration example hard-codes a Chinese name and Chinese payload message, which steers invocation into a specific language/locale. The file does not state that the skill is region-specific or provide an opt-in or alternative language option, so this is a natural-language policy issue under the locale/language rule.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.