Back to skill

Security audit

Brand Protection Walmart

Security checks for vulnerabilities and agentic risk

Overview

The skill is advertised as Walmart brand protection, but its shipped scripts generate Amazon-focused detection reports and complaint templates.

Review this carefully before installing. The included tools appear to be local template/report generators, but they are not reliably Walmart-specific despite the advertised name and description. Do not use the generated enforcement guidance for Walmart cases unless the marketplace language, portal links, identifiers, and workflows are corrected. Prefer a pinned, local install instead of the documented global unpinned npx command.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:16
Finding
Unpinned Third-Party Package Execution During Global Installation## Vulnerability Details **File Location**: `SKILL.md:16` **Vulnerability Type**: Supply-chain risk from an unpinned third-party installer **Risk Level**: Medium ### Vulnerable Code ```bash npx skills add nexscope-ai/eCommerce-Skills --skill brand-protection-walmart -g ``` ### Technical Analysis The documented installation procedure invokes the remotely resolved `skills` npm package through `npx` without specifying an exact package version or integrity hash. It then installs content from the mutable `nexscope-ai/eCommerce-Skills` source globally. Because neither dependency is pinned to an immutable, verified revision, the code executed or installed at audit time may differ from the code later received by users. An attacker who compromises the npm package, its maintainer account, or the referenced repository could introduce arbitrary installation-time behavior after this reviewed project has been published. The audited local Python scripts do not themselves perform remote retrieval or command execution. The risk arises specifically when a user follows the installation instruction. ### Attack Path 1. An attacker compromises the npm package, package-maintainer account, or referenced external repository. 2. The attacker publishes or commits a modified installer or skill payload. 3. A user follows the documented command without specifying a trusted version or commit. 4. `npx` resolves and executes the current remote package. 5. The installer retrieves mutable repository content and installs it globally. 6. Malicious package lifecycle code, installer logic, or installed skill content executes with the invoking user's privileges. ### Impact Assessment Successful exploitation could permit arbitrary code execution under the account running the installation command. Depending on that account's permissions and the installer's behavior, the attacker could access user-readable files, modify globally managed skill content, tamper with development configuration, or ...[truncated 289 chars]
Remediation
## Remediation Suggestions 1. Pin the npm package to a reviewed exact version rather than relying on the latest remotely resolved release. 2. Pin the external repository to an immutable commit hash or signed release tag. 3. Publish and verify cryptographic integrity information for downloaded artifacts. 4. Avoid global installation by default; prefer a project-local, least-privilege installation. 5. Review package lifecycle scripts and the complete transitive dependency tree before recommending execution. 6. Use lockfiles or an equivalent reproducible dependency manifest. 7. Require signed releases and protected maintainer accounts with multi-factor authentication. 8. Document a verification procedure that allows users to confirm the package version, repository revision, and checksums before installation.
Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (11)

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding
A marketplace mismatch—claiming Walmart support while actually targeting Amazon workflows or only producing generic templates—creates deceptive security functionality. In the context of a brand-protection skill, this can cause missed infringement response windows, improper reporting, and operational decisions based on false assumptions that active detection or Walmart portal integration is in place.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding
A marketplace mismatch—claiming Walmart support while actually targeting Amazon workflows or only producing generic templates—creates deceptive security functionality. In the context of a brand-protection skill, this can cause missed infringement response windows, improper reporting, and operational decisions based on false assumptions that active detection or Walmart portal integration is in place.

Description-Behavior Mismatch

High
Confidence
98% confidence
Finding
The implementation and generated action plan are explicitly Amazon-focused, recommending Amazon Brand Registry, Amazon counterfeit complaints, and ASIN-based workflows despite the skill being ներկայացված as Walmart-specific. This can mislead operators into taking the wrong enforcement actions, causing failed incident response, delayed brand protection, and potential disclosure of sensitive case details to the wrong platform or process.

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding
npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Description-Behavior Mismatch

Medium
Confidence
93% confidence
Finding
The core data model is built around Amazon-specific identifiers and controls such as ASINs and Brand Registry rather than Walmart-native entities. In a brand-protection skill, this schema mismatch can propagate incorrect assumptions through detection, reporting, and operator decisions, increasing the chance of ineffective or misdirected enforcement actions.

Natural-Language Policy Violations

Medium
Confidence
92% confidence
Finding
The CLI selects English by default, but the report formatter contains a built-in branch that produces Chinese-specific output when `lang == "zh"`, and the program enables this locale based solely on a command-line flag rather than an interactive user language choice. This creates a locale-forcing behavior path in the skill logic without any broader language preference negotiation or explicit user opt-in inside the skill itself.

Description-Behavior Mismatch

Medium
Confidence
97% confidence
Finding
The manifest describes a Walmart brand protection toolkit with Walmart Brand Portal reporting and WFS seller monitoring, but this file generates complaint content explicitly for Amazon Brand Registry and links to Amazon's portal. That is a direct mismatch between the declared platform-specific purpose and the actual behavior implemented here.

Description-Behavior Mismatch

Medium
Confidence
94% confidence
Finding
The letter text states the conduct occurs on Amazon.com and threatens reporting to Amazon Brand Registry, which does not align with a Walmart-specific toolkit description. This is not a generic implementation detail; it changes the marketplace and enforcement channel the skill actually supports.

Description-Behavior Mismatch

Medium
Confidence
93% confidence
Finding
The guide instructs users to create a separate Amazon buyer account and later submit complaints via Brand Registry, which conflicts with the manifest's Walmart-oriented scope. The implemented content therefore supports a different marketplace than the one advertised.

Natural-Language Policy Violations

Medium
Confidence
90% confidence
Finding
This file includes a locale switch that forces Chinese output whenever `--zh` is present, and the file provides no natural-language justification that the skill is intended only for a Chinese-speaking or region-specific context. The policy requires either explicit user language choice/opt-in in the skill behavior or a clearly documented locale constraint.

Description-Behavior Mismatch

Low
Confidence
87% confidence
Finding
The MAP notice includes Amazon-specific escalation language even though the skill is described as a Walmart brand protection toolkit. While MAP enforcement is within brand protection generally, the hardcoded Amazon escalation path diverges from the claimed Walmart platform focus.

Static analysis

No suspicious patterns detected.