Back to skill

Security audit

Maxhub Pipixia

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill suspicious because its references describe capabilities beyond the stated Pipixia data-query purpose, and its installation command is unpinned; no clear malicious intent is reported.

Consider avoiding installation until the publisher explains the extra capabilities and provides a version-pinned installation method. The scan inspected all seven listed components, but analysis was marked partial because some references could not be resolved, and the installed package implementation was not included; 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (8)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
92% confidence
Finding

The only supplied executable content is an installation command, so its demonstrated behavior differs from the declared query-assistant purpose. It may be setup documentation for a matching skill, but the installed package's implementation is not provided and cannot be verified.

Content

No source excerpt is available for this finding.

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

Medium
Category
MCP Rug Pull
Confidence
92% confidence
Finding

The installation command invokes an unpinned npm executable, allowing the code executed during installation to change without changes to this documentation. Although clawhub is an installer rather than an MCP server, a compromised future package release could execute code with the installing user's permissions.

Content

No source excerpt is available for this finding.

Example invocation prompts are generic and do not specify the PiPiXia context.

Medium
Category
Trigger Abuse
Confidence
94% confidence
Finding

The example prompts include 用户 (user), 搜索 (search), and 浏览 (browse), which overlap with everyday requests unrelated to PiPiXia. Although the README identifies the platform, these invocation examples provide no platform qualification or exclusion conditions to prevent unintended activation.

Content

No source excerpt is available for this finding.

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

Medium
Category
MCP Rug Pull
Confidence
92% confidence
Finding

The documented npx command does not pin the clawhub executable, so installation can execute a different package release over time. The MCP-server label is inaccurate, but the underlying supply-chain risk remains: a compromised release could run with the installer's local permissions.

Content

No source excerpt is available for this finding.

Usage examples present generic words such as “用户” (user), “搜索” (search), and “浏览” (browse) as invocation instructions without platform-specific qualifiers.

Medium
Category
Trigger Abuse
Confidence
90% confidence
Finding

These example instructions overlap with ordinary requests unrelated to Pipixia. Although the introduction identifies the skill's domain, the invocation examples do not require that context or explain when generic requests should not activate the skill.

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

# POST example

The API reference includes a view-count mutation despite the manifest describing a data-query assistant.

Medium
Category
Excessive Agency
Confidence
96% confidence
Finding

The reference embeds parameters and a stated purpose for increasing post views among otherwise read-only query documentation, creating a risk of unauthorized engagement modification. This contradicts the skill's read-only assurances, which are not evidence of safety; however, the missing endpoint path limits evidence that this fragment alone is executable.

Content

No source excerpt is available for this finding.

The skill exposes arbitrary-URL shortening, which is unrelated to querying Pipixia data.

Low
Category
Excessive Agency
Confidence
94% confidence
Finding

The documented fetch_short_url endpoint generates short links, and L135 explicitly permits arbitrary input URLs. This general-purpose link-generation capability is not an obvious requirement of the manifest's stated purpose of querying Pipixia content, users, search, trends, comments, and topics.

Content

No source excerpt is available for this finding.