Back to skill

Security audit

卖家精灵-市场调研(列表)

Security checks across malware telemetry and agentic risk

Overview

The skill has a real SellerSprite market-research purpose, but it adds credential-sensitive routing, mandatory local retention, automatic feedback reporting, and a runtime dependency-install path that users should review.

Install only if you trust LinkFox/SellerSprite and are comfortable with paid-credit API calls and full market-research responses being stored locally. Keep LINKFOX_TOOL_GATEWAY unset unless you control the destination, review or disable automatic feedback behavior, and do not allow the skill to install the onboarding package unless you intentionally want that extra skill added.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Findings (8)

Tainted flow: 'req' from os.environ.get (line 72, credential/environment) → urllib.request.urlopen (network output)

Critical
Category
Data Flow
Content
method="POST",
    )
    try:
        with urlopen(req, timeout=120) as response:
            return json.loads(response.read().decode("utf-8"))
    except HTTPError as e:
        body = e.read().decode("utf-8") if e.fp else ""
Confidence
96% confidence
Finding
with urlopen(req, timeout=120) as response:

Lp3

Medium
Category
MCP Least Privilege
Confidence
88% confidence
Finding
The skill describes capabilities to read environment variables, write files, and call network resources, yet no explicit permissions are declared. This creates a transparency and policy-enforcement gap: operators and users cannot accurately assess what the skill may access, and hidden capability expansion is especially risky because the skill also handles API keys, downloads, and persistent logging.

Context-Inappropriate Capability

Medium
Confidence
93% confidence
Finding
The skill instructs automatic submission of feedback about user behavior and skill performance to a separate Feedback API, even though that data flow is not necessary for core market-research functionality. This creates an unjustified secondary exfiltration path for user interaction metadata and potentially sensitive intent signals without clear consent or minimization.

Context-Inappropriate Capability

High
Confidence
97% confidence
Finding
The skill directs the agent to download and install another skill package from an external URL when onboarding is missing, which materially exceeds the stated market-research purpose. This introduces a supply-chain and privilege-escalation risk because unreviewed remote content can change agent behavior, add new capabilities, or execute additional instructions under the guise of troubleshooting.

Vague Triggers

High
Confidence
86% confidence
Finding
The trigger conditions are overly broad, causing the skill to activate for generic Amazon market-research requests even when SellerSprite is not mentioned. Overbroad activation can route unrelated user requests into a tool that consumes paid credits, accesses external services, persists data locally, and may apply hidden operational rules the user did not ask for.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill mandates persistent local storage of full API responses in a session-organized path without a clear up-front user warning or consent flow. Full responses can contain user-supplied search criteria, account-related metadata, or other sensitive business data, so mandatory retention increases exposure beyond what is necessary to answer the immediate request.

Missing User Warnings

Medium
Confidence
75% confidence
Finding
The curl example demonstrates placing the Authorization header directly on the command line, which can expose secrets through shell history, process inspection, shared terminals, CI logs, or support screenshots. In a skill integration context, examples are often copied verbatim, so insecure example usage can directly lead to credential leakage.

Ssd 3

Medium
Confidence
96% confidence
Finding
The skill requires unconditional logging of complete API responses to local files organized by date and session, which creates unnecessary retention of potentially sensitive business inputs and outputs. Session-based accumulation magnifies risk by building a longitudinal record that may be accessible to other local processes, future tasks, or operators beyond the immediate user interaction.

VirusTotal

60/60 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.