Back to skill

Security audit

中国效率工具集

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent China-focused lookup toolkit, but it sends personal identifiers to third-party APIs and uses plaintext HTTP for IP lookups without clear user-facing disclosure.

Review this skill before installing if you may query real phone numbers, IP addresses, or package tracking numbers. Those values are sent to public third-party services, and IP lookups currently use unencrypted HTTP; avoid submitting sensitive identifiers unless that exposure is acceptable.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (1)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/toolkit.js:129
Finding

Plaintext HTTP Used for IP Geolocation Requests

Content
View full analysis

Vulnerability Details

File Location: scripts/toolkit.js:129
Vulnerability Type: Unencrypted transmission of user input and API responses
Risk Level: Medium

Vulnerable Code

js
const data = await fetch(`http://ip-api.com/json/${ip}?lang=zh-CN`);

Technical Analysis

The IP geolocation function sends the user-supplied IP address to ip-api.com over plaintext HTTP. HTTP provides neither transport confidentiality nor server authentication and response integrity.

A passive network observer can read the queried IP address and returned geolocation information. An active on-path attacker can intercept the request and replace the response with attacker-controlled JSON. Because the application accepts parsed JSON without cryptographic verification, a forged response with status: "success" is treated as legitimate and its country, region, city, ISP, organization, and timezone values are printed to the caller.

Attack Path

  1. A user invokes china-toolkit ip <IP-address>.
  2. The skill constructs a plaintext request to http://ip-api.com/json/....
  3. An attacker with an on-path position, such as control of an untrusted Wi-Fi access point, compromised router, or network proxy, intercepts the request.
  4. The attacker reads the queried IP address or returns forged JSON containing status: "success" and manipulated geolocation fields.
  5. The skill parses the response and presents the forged information as successful API output.

Impact Assessment

Exploitation does not grant local code execution, filesystem access, elevated system privileges, or persistence. Its scope is limited to the confidentiality of IP lookup queries and the integrity of geolocation results transmitted through this endpoint. Manipulated results could cause a user or downstream agent to make incorrect decisions based on forged country, organization, ISP, or location data.

Remediation
View remediation

Remediation Suggestions

  • Replace the endpoint with an IP geolocation service that supports HTTPS.
  • Do not silently fall back or redirect to plaintext HTTP.
  • Retain Node.js TLS certificate and hostname verification.
  • Validate response status codes, content type, structure, and expected field types before using the response.
  • Add request timeouts and response-size limits to reduce exposure to stalled or oversized third-party responses.
  • Document that IP addresses are disclosed to a third-party geolocation provider.
  • Add a regression test that rejects non-HTTPS service URLs.
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 (7)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill explicitly states it uses public third-party APIs and includes phone-number and IP lookup features, but it does not warn users that these queries transmit potentially sensitive data off-platform. This creates a privacy and data-handling risk because users may submit personal identifiers without informed consent, and public API providers may log, retain, or further process that information.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The phone lookup function transmits a user-supplied phone number to a third-party endpoint without any disclosure, consent flow, or minimization. Phone numbers are personal data, and sending them to an external service can expose sensitive user information, create compliance/privacy issues, and leak data to an unvetted provider.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The IP lookup feature sends user-supplied IP addresses to a third-party geolocation service without warning or consent. IP addresses can be personal data, and the implementation is more concerning because it uses plain HTTP, allowing interception or tampering of the queried IP and returned geolocation data in transit.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

The skill is presented entirely as a Chinese-environment toolkit and the documentation is fully in Chinese, with no indication that users can opt into another language or locale. The policy requires flagging language or locale constraints when they are imposed without user choice or explicit justification.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

This manifest description lists the skill's capabilities in broad terms but provides no explicit trigger phrases, scope limits, or exclusion conditions. For manifest files, that can make invocation criteria ambiguous and increase the chance of unintended activation based on generic references to these common tasks.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
87% confidence
Finding

The file presents the toolkit as China-focused and all user-facing help and error strings are in Chinese, with no indication of language choice or opt-in. This can constitute a language/locale policy issue because the skill effectively forces a specific language experience on users.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The request explicitly sets lang=zh-CN, forcing Chinese-language localization for the returned data without any user opt-in or configuration. This is a natural-language policy concern because the locale is imposed by default rather than chosen by the user.

Content

No source excerpt is available for this finding.