Back to skill

Security audit

parcel-coupon

Security checks for vulnerabilities and agentic risk

Overview

The skill performs the advertised coupon lookup, but it persists and transmits a unique install identifier for attribution without clear in-skill controls.

Review this skill before installing if you do not want coupon providers or the API operator to correlate requests across runs. The coupon links and QR codes are external promotional destinations, so verify domains before opening links or scanning codes, and prefer explicit coupon requests rather than allowing broad shipping questions to trigger the skill.

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 (8)

Rp1

Medium
Category
MCP Rug Pull
Confidence
89% confidence
Finding

The README instructs users to install the skill via npx skills add moooai/parcel-coupon without pinning a specific version of the installer/runtime. This can lead to non-reproducible installs and creates a supply-chain risk if the referenced package or its transitive dependencies are updated maliciously or unexpectedly.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The suggested trigger phrases such as asking whether there are shipping coupons are broad, natural-language requests that a host assistant could match in ordinary conversation. This increases the chance of unintended invocation of the skill, which is more concerning here because the skill promotes external coupon links and tracking-related attribution parameters (ch/u).

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
81% confidence
Finding

The skill repeatedly instructs users to open links in WeChat or scan QR codes in WeChat, strongly steering them into a specific external platform without presenting alternatives or clearly justifying the requirement. In a security context, this reduces user choice and can make phishing, tracking, or opaque cross-platform redirection harder for users to evaluate before leaving the assistant environment.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The trigger phrases are broad enough to match generic shipping-help or savings queries, which can cause the skill to activate outside a clearly bounded coupon-only context. That increases the chance of unsolicited promotion, accidental network calls, and steering users toward third-party links/QR codes when they did not explicitly ask for coupon redemption.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The docstring describes the script as only fetching coupon information, but the implementation also creates a persistent per-installation UUID and transmits it on every request for attribution tracking. This hidden behavior undermines transparency and informed consent, and can enable cross-run tracking of users or hosts by the remote service.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The code generates a stable install ID, stores it locally, and automatically sends it to a remote API as the 'u' parameter without any notice or consent flow. Even if not directly sensitive by itself, a persistent unique identifier enables user/device correlation over time and is especially concerning because it is unnecessary for the advertised core function of simply retrieving coupons.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

The entire API documentation is presented only in Chinese, including operational guidance and error messages, with no indication that the skill is region-specific or that users may opt into another language. This can violate language/locale policy when a skill imposes a specific language without user choice or explicit justification.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
85% confidence
Finding

The output strings instruct users to open the coupon link with WeChat and scan the QR code with WeChat, which imposes a specific platform/language ecosystem constraint in natural language. The file does not present this as an optional preference or explain why this restriction is required.

Content

No source excerpt is available for this finding.