Back to skill

Security audit

Maxhub Xigua

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill suspicious because its video-query fallbacks can send requests to Douyin endpoints outside its stated Xigua scope; the scan evidence is partly incomplete.

Consider the platform-switching behavior and unpinned installation command before installing. The scanner reported no executable scripts, but its analysis was marked partial because two local references were unresolved, so the evidence is incomplete.

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
  • 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
Findings (6)

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

Medium
Category
MCP Rug Pull
Confidence
88% confidence
Finding

The installation command invokes an unpinned npm CLI through npx, so a fresh installation can execute a different package release without review. Although the finding incorrectly labels this as an MCP server, the underlying supply-chain risk is real: a compromised release could execute code with the installing user's permissions.

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 installation example uses npx without pinning the clawhub package version, allowing the executed release to change between installations. This is not an MCP server reference, but it does introduce a supply-chain exposure if a subsequently resolved release is compromised.

Content

No source excerpt is available for this finding.

Example invocation phrases are overly broad: “视频,用户” (“video, user”).

Medium
Category
Trigger Abuse
Confidence
91% confidence
Finding

The usage table presents generic nouns as example commands without specifying a Xigua Video lookup or another concrete action. Although the introduction identifies the platform, these invocation examples overlap with ordinary conversation and do not clarify when the skill should activate.

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/xigua/{endpoint}?%7Bparams=}" \
  -H "$maxhub_auth_header"

# POST example

Xigua video-query fallbacks route requests to Douyin endpoints outside the stated platform scope.

Medium
Category
Excessive Agency
Confidence
97% confidence
Finding

The automatic fallback table directs a Xigua-only assistant to Douyin endpoints without obtaining user approval for a platform change. This can submit query identifiers to unintended platform-specific processing, consume API quota, and produce misleading cross-platform results; the destination remains the declared MaxHub provider, so this is not evidence of exfiltration to a new host.

Content

No source excerpt is available for this finding.

Example invocation prompts are generic nouns that do not specify Xigua or a concrete task.

Low
Category
Trigger Abuse
Confidence
88% confidence
Finding

The example prompts “视频” (video) and “用户” (user) are common terms that could refer to many unrelated tasks. Although the README identifies the skill as Xigua-specific, these invocation examples do not express that constraint or explain when the skill should not activate.

Content

No source excerpt is available for this finding.