Back to skill

Security audit

Maxhub Linkedin

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marked this skill suspicious because its instructions raise concerns about use beyond LinkedIn and disclosure of search data, alongside an unpinned installation command; the evidence does not establish malicious intent.

Review the skill’s LinkedIn-only scope, its Douyin fallback, and what search terms or identifiers it sends to MaxHub before installing. The installer command is unpinned, and the scan was partial because two references could not be resolved, so the available evidence does not provide a complete review.

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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (7)

MCP server referenced without pinned version: 'npx clawhub'.

Medium
Category
MCP Rug Pull
Confidence
88% confidence
Finding

The installation command executes an unpinned npm package through npx, so a future compromised release could execute code with the installing user's privileges. This is a supply-chain hardening issue, although the finding incorrectly describes the CLI as an MCP server.

Content

No source excerpt is available for this finding.

The example invocation prompts use generic words without requiring LinkedIn-specific context.

Medium
Category
Trigger Abuse
Confidence
96% confidence
Finding

The usage table presents standalone terms such as 用户 (user), 搜索 (search), and 评论 (comments) as example prompts. These overlap with everyday requests across many domains, and the README does not explicitly restrict activation to LinkedIn-related requests or provide exclusion conditions.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx clawhub'.

Medium
Category
MCP Rug Pull
Confidence
88% confidence
Finding

The documented npx command executes a CLI without pinning its package version, exposing installers to malicious changes in a future resolved release. Execution occurs with the user's privileges; the relevant risk is package supply-chain compromise, not an unpinned MCP server.

Content

No source excerpt is available for this finding.

Invocation examples use generic words without LinkedIn-specific scope.

Medium
Category
Trigger Abuse
Confidence
96% confidence
Finding

The example commands include everyday terms such as “用户” (user), “公司” (company), and “搜索” (search). Although the README describes a LinkedIn assistant, these invocation examples lack LinkedIn qualifiers or exclusion conditions and could cause unintended activation for unrelated requests.

Content

No source excerpt is available for this finding.

The workflow sends query parameters to MaxHub without warning users that their search terms and queried identifiers will be shared with a third-party service.

Medium
Category
Output Handling
Confidence
88% confidence
Finding

The skill instructs the agent to query MaxHub and execute planned queries autonomously, but provides no privacy warning about transmitting user-supplied search terms or profile/company identifiers. Naming the API provider and disclaiming data accuracy do not explain this disclosure of query data; read-only requests can still expose sensitive research interests.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% 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 53)May include surrounding context.

md
maxhub_auth_header="Authorization: Bearer ${MAXHUB_API_KEY}"

# GET example
curl -s "https://www.aconfig.cn/api/v1/linkedin/{endpoint}?%7Bparams=}" \
  -H "$maxhub_auth_header"

# POST example

The fallback workflow prescribes Douyin API calls outside the declared LinkedIn-only scope.

Medium
Category
Excessive Agency
Confidence
95% confidence
Finding

The manifest at L003 describes a LinkedIn data-query assistant, but the operational fallback table specifies requests to Douyin video, search, and profile endpoints. These are not merely illustrative URLs: L293 instructs the agent to use this table when requests fail, introducing cross-platform behavior unrelated to the declared purpose.

Content

No source excerpt is available for this finding.