Back to skill

Security audit

Lafeitu (辣匪兔) Spicy Food

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed Lafeitu shopping skill that can manage account, cart, profile, and unpaid order actions through the official API when directed by the user.

Install only if you want an agent to interact with your Lafeitu account and shopping cart. Confirm product choices, quantities, profile changes, recipient phone number, and shipping address before running mutating commands, and use logout or delete the credential file when you no longer want the saved token retained.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Rogue AgentSelf-Modification, Session Persistence
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (8)

Lp3

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding
The skill invokes python scripts, performs network access to a remote API, and stores data under a local credentials path, but it does not declare corresponding permissions or clearly constrain those capabilities. This creates a trust and review gap: users and platforms may approve the skill without realizing it can access the network, shell, and local files, increasing the chance of unintended data access or command execution.

Description-Behavior Mismatch

Medium
Confidence
92% confidence
Finding
The client is explicitly designed as a universal multi-merchant engine, which conflicts with the skill's declared brand-specific scope for Lafeitu. In an agent setting, this scope expansion can enable the skill to be repurposed to interact with unrelated merchants, violating least-privilege expectations and increasing the risk of cross-merchant data handling or unauthorized commerce actions.

Description-Behavior Mismatch

High
Confidence
98% confidence
Finding
The constructor accepts any HTTPS endpoint except non-localhost plain HTTP, so the skill can send authenticated commerce requests and persist credentials for arbitrary domains rather than only the official lafeitu.cn API promised in the manifest. This creates a real trust-boundary violation: a malicious or misconfigured caller could direct the client to an attacker-controlled HTTPS host and exfiltrate account identifiers, tokens, cart data, profile data, or shipping information.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The README documents account-bound and state-changing operations such as login, registration, profile updates, cart modification, and order creation without clearly warning that these actions transmit user data to a remote third-party service and may alter account or order state. In an agent skill context, unclear disclosure increases the chance that an agent or user triggers sensitive operations without informed consent, especially when credentials and personal profile data are involved.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill states that credentials are stored locally under a persistent path, yet it provides no explicit privacy notice, retention guidance, or user confirmation before saving tokens/profile-related data. Local persistence of authentication material raises the risk of credential theft, account misuse, or unintended cross-session access on shared systems.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The order flow collects and transmits sensitive personal data including name, phone number, province, city, and street address, but the skill does not explicitly warn users about this transmission or describe handling safeguards. This increases privacy risk and may lead users to disclose personal information without informed consent or understanding of where it is sent.

Session Persistence

Medium
Category
Rogue Agent
Content
# Lafeitu Gourmet Skill

A brand-specific commerce skill for Lafeitu (辣匪兔). This project gives agents a structured way to browse products, manage carts, handle account flows, and create orders against the official `https://lafeitu.cn/api/v1` backend, while leaving final payment to the user.

Official Website: https://lafeitu.cn
clawhub: https://clawhub.com/NowLoadY/agentic-spicy-food
Confidence
68% confidence
Finding
create orders against the official `https://lafeitu.cn/api/v1` backend, while leaving final payment to the user. Official Website: https://lafeitu.cn clawhub: https://clawhub.com/NowLoadY/agentic-spi

Session Persistence

Medium
Category
Rogue Agent
Content
- inspect variants, pricing, promotions, or shipping thresholds
- manage a Lafeitu cart
- log in, register, or update Lafeitu profile data
- create a Lafeitu order and hand payment back to the user
- retrieve official Lafeitu brand, company, or contact information

## Tool Priority & Fallback Strategy
Confidence
77% confidence
Finding
create a Lafeitu order and hand payment back to the user - retrieve official Lafeitu brand, company, or contact information ## Tool Priority & Fallback Strategy 1. **API first**: Use `python3 script

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.