Back to skill

Security audit

Brand Protection Tiktok

Security checks for vulnerabilities and agentic risk

Overview

The skill claims to protect brands on TikTok Shop, but its actual scripts generate Amazon-focused detection output, complaint templates, and enforcement guidance.

Review this skill carefully before installing. It should not be relied on for TikTok Shop enforcement as packaged, because its evidence models and legal/reporting templates point users toward Amazon workflows. If installed, prefer an isolated or project-local install path and pin or verify the installer version rather than running the unpinned global 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 Package Execution During Global Installation## Vulnerability Details **File Location**: `SKILL.md:16` **Vulnerability Type**: Unpinned third-party dependency execution **Risk Level**: Medium ### Vulnerable Code ```bash npx skills add nexscope-ai/eCommerce-Skills --skill brand-protection-tiktok -g ``` ### Technical Analysis The documented installation procedure invokes the `skills` npm package through `npx` without specifying an exact version or integrity hash. Consequently, the package resolved and executed at installation time may differ from the version that existed when this Skill was audited. Because `npx` may download and execute package code, including lifecycle behavior, compromise of the resolved package, its dependencies, or the relevant package distribution account could introduce arbitrary code into the installation process. The `-g` option also requests global Skill installation, increasing the persistence and scope of unauthorized modifications within the user's Skill environment. The repository's local Python scripts do not themselves retrieve remote code, execute subprocesses, access credentials, write files, establish persistence, or exfiltrate data. The risk arises specifically from the documented external installation command. ### Attack Path 1. An attacker compromises the npm package resolved as `skills`, one of its transitive dependencies, or its publishing account. 2. The attacker publishes a malicious package version containing installation-time or runtime code. 3. A user follows the command documented in `SKILL.md`. 4. `npx` resolves and downloads the current unpinned package version. 5. The malicious package code executes with the privileges of the user running the installation. 6. The package can alter the globally installed Skill environment or perform any operation available to that user. ### Impact Assessment Successful exploitation could provide arbitrary code execution with the installing user's privileges. Depending on that user's permissions, an attacker could re ...[truncated 465 chars]
Remediation
## Remediation Suggestions 1. Pin the installer to an exact, reviewed version rather than relying on the latest package resolution: ```bash npx --yes skills@X.Y.Z add nexscope-ai/eCommerce-Skills --skill brand-protection-tiktok ``` 2. Verify the selected package version and all transitive dependencies through a committed lockfile and package integrity metadata. 3. Publish and document expected package provenance, registry, maintainer identity, and cryptographic checksums or signatures. 4. Avoid global installation unless it is necessary. Prefer a project-local or isolated environment to reduce the blast radius. 5. Review lifecycle scripts and dependency changes before upgrading the pinned installer. 6. In security-sensitive deployments, obtain the installer from a trusted internal mirror and enforce package allowlisting and integrity verification.
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • 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
Findings (11)

Tp4

High
Category
MCP Tool Poisoning
Confidence
93% confidence
Finding
This is a second description-behavior mismatch report for the same location and underlying issue: the skill presents itself as a TikTok brand-protection toolkit while apparently only offering generic or non-TikTok-targeted template functionality. In a security-sensitive and compliance-oriented context, misleading capability claims can cause operational blind spots, wasted incident response time, and potentially improper reporting actions.

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding
This is a second description-behavior mismatch report for the same location and underlying issue: the skill presents itself as a TikTok brand-protection toolkit while apparently only offering generic or non-TikTok-targeted template functionality. In a security-sensitive and compliance-oriented context, misleading capability claims can cause operational blind spots, wasted incident response time, and potentially improper reporting actions.

Description-Behavior Mismatch

High
Confidence
99% confidence
Finding
The skill is marketed as a TikTok Shop brand-protection toolkit, but the implementation is Amazon-specific throughout, using ASINs, Brand Registry, Amazon complaint flows, and Amazon-focused recommendations. This is dangerous because users may rely on it for TikTok enforcement decisions and generate incorrect reports or miss actual platform-specific abuse, causing operational, legal, and trust harm.

Description-Behavior Mismatch

High
Confidence
99% confidence
Finding
The skill is advertised as a TikTok Shop brand-protection toolkit, but the generated complaint template directs users to Amazon Brand Registry instead. In a legal-enforcement workflow, this mismatch can cause users to submit complaints through the wrong platform, mishandle evidence, and rely on inapplicable procedures, making the skill materially unsafe in context.

Description-Behavior Mismatch

High
Confidence
97% confidence
Finding
The cease-and-desist letter is framed around Amazon.com enforcement and Amazon anti-counterfeiting policy, not TikTok Shop. Because this skill is positioned as a TikTok brand-protection tool, users may send legally and operationally incorrect notices, weakening enforcement efforts and creating reputational or legal risk through inaccurate platform references.

Description-Behavior Mismatch

High
Confidence
98% confidence
Finding
The test-buy guide gives Amazon-specific account, purchasing, and Brand Registry complaint steps rather than TikTok Shop procedures. In this skill context, that can mislead investigators into following the wrong evidence-collection and reporting path, reducing effectiveness and potentially causing account-policy or procedural violations.

Rp1

Medium
Category
MCP Rug Pull
Confidence
84% confidence
Finding
The installation command uses `npx skills` without pinning a specific version, which can cause users to execute an unexpected or compromised package version at install time. Because `npx` fetches and runs code dynamically, a supply-chain compromise or breaking upstream change could affect anyone following the documented command.

Description-Behavior Mismatch

Medium
Confidence
93% confidence
Finding
The manifest claims influencer misuse detection and complaint-template functionality, but this file only performs seller/listing heuristics and report formatting. That mismatch can mislead users into assuming important enforcement capabilities exist when they do not, leading to incomplete investigations and delayed response to abuse.

Description-Behavior Mismatch

Medium
Confidence
92% confidence
Finding
The MAP notice threatens escalation by reporting to Amazon, which is inconsistent with the stated TikTok Shop focus. While less severe than the core complaint templates, it still provides misleading enforcement guidance and could cause users to invoke the wrong escalation path in a commercial dispute.

Natural-Language Policy Violations

Low
Confidence
87% confidence
Finding
This code hard-codes the report language to English unless the operator knows to supply the `--zh` flag. That creates a natural-language locale preference in the skill behavior without an explicit user opt-in mechanism or documented justification for the default constraint.

Natural-Language Policy Violations

Low
Confidence
85% confidence
Finding
This code selects output language based solely on the presence of a --zh flag and otherwise defaults to English, while also exposing separate Chinese template generators. There is no natural-language notice explaining language behavior or offering broader locale choice, which can conflict with policies requiring user opt-in or clear language selection.

Static analysis

No suspicious patterns detected.