Back to skill

Security audit

1688 商品详情

Security checks across malware telemetry and agentic risk

Overview

The skill mainly performs a 1688 product lookup, but it also bundles account login, billing/payment, credential output, local retention, and automatic feedback reporting that users should review carefully.

Install only if you are comfortable with LinkFox receiving product lookup requests, API credentials, session/app metadata, and optional feedback content, and if you accept that auth and billing support are bundled into this lookup skill. Avoid sending sensitive conversation text through feedback, protect generated API keys, and review or clean the local linkfox data/cache directories if supplier or pricing data is sensitive.

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 (17)

Tainted flow: 'req' from os.environ.get (line 70, 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
95% confidence
Finding
with urlopen(req, timeout=120) as response:

Lp3

Medium
Category
MCP Least Privilege
Confidence
94% confidence
Finding
The skill uses environment variables, network access, and file writes but does not declare these capabilities or permissions. Hidden capabilities reduce transparency and reviewability, and in this skill the file-write and env access are especially relevant because responses are persisted to disk and API keys are read from the environment.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding
The documented purpose is a product-detail lookup, but the referenced behavior reportedly includes account login, SMS verification, API-key generation, plan listing, payment order creation, QR-code payment, and order-status checks. That is a major trust-boundary violation: a user invoking a read-only sourcing lookup could unintentionally trigger authentication or payment-related flows with financial and account-security consequences.

Context-Inappropriate Capability

Medium
Confidence
87% confidence
Finding
The skill instructs automatic feedback reporting for broad conditions, including user praise/dissatisfaction and anything that could be improved, which is unrelated to the narrow product-detail function. Automatic background reporting can exfiltrate user interaction data or metadata to an external API without clear user awareness or necessity.

Description-Behavior Mismatch

High
Confidence
97% confidence
Finding
The file adds authentication recovery, account onboarding, and billing/payment workflows to a skill whose declared purpose is only retrieving 1688 product details. This scope expansion creates an unnecessary path for handling credentials and account lifecycle actions, increasing the chance of abuse, phishing-style collection, or accidental execution of unrelated sensitive operations.

Context-Inappropriate Capability

High
Confidence
96% confidence
Finding
The documentation instructs the agent to facilitate registration, key acquisition, plan listing, order creation, and payment follow-up, none of which are justified by a sourcing-detail use case. Introducing these capabilities enables collection of sensitive user data and transitions the skill into account and payment operations, which materially raises security and fraud risk.

Description-Behavior Mismatch

High
Confidence
99% confidence
Finding
This file implements account onboarding, SMS authentication, API-key retrieval, subscription browsing, order creation, and payment QR generation, which is materially unrelated to a 1688 product-detail lookup skill. In this context, the mismatch is dangerous because it expands the skill from data lookup into credential handling and billing operations, increasing the blast radius if invoked or misused.

Context-Inappropriate Capability

High
Confidence
98% confidence
Finding
Embedding billing and payment creation inside a sourcing-detail skill introduces the ability to create orders and generate payment artifacts without being justified by the declared purpose. In an agent setting, this can enable unauthorized or confusing commercial actions under the guise of a harmless lookup capability.

Context-Inappropriate Capability

High
Confidence
99% confidence
Finding
The code performs SMS login and then generates or fetches API tokens, exposing a credential-issuance workflow that is not needed for 1688 product-detail lookup. In skill context, this is especially dangerous because users may invoke it expecting product data while the tool actually handles authentication secrets and account access.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The Feedback API section instructs clients to POST free-form `content` to `https://skill-api.linkfox.com` without clearly warning that the text is transmitted to an external service. In a skill context, feedback content can easily include user prompts, product identifiers, supplier details, or other sensitive conversation data, so omission of an explicit disclosure increases the risk of unintended data exfiltration and privacy/compliance violations.

Vague Triggers

Medium
Confidence
84% confidence
Finding
The activation logic keys off broad error strings such as 'auth', 'billing', 'unauthorized', and similar text, which can appear in unrelated contexts or be induced by adversarial input. That can cause the agent to pivot from benign product lookup into sensitive onboarding or payment flows without adequate scope verification.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill instructs collection of a user's phone number for scripted registration but provides no privacy notice, retention policy, or consent language. Collecting personal data in an unrelated product-detail skill increases the risk of improper handling, user deception, and unnecessary exposure of sensitive identifiers.

Missing User Warnings

Medium
Confidence
86% confidence
Finding
The script transmits SESSION_ID, MODE_ID, and APP_NAME from the environment to a remote service without any explicit disclosure or consent flow in the script itself. These identifiers may contain user, tenant, or workflow metadata, so silently forwarding them expands data exposure and can aid correlation or tracking across sessions.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The script persistently stores the full API response locally by default, which may include supplier, pricing, SKU, logistics, or other sensitive procurement data, without presenting an explicit privacy/safety warning at runtime. In agent environments and shared workspaces, silent persistence increases the chance of unintended retention, later disclosure, or cross-task data leakage.

Missing User Warnings

High
Confidence
98% confidence
Finding
The login command emits the retrieved API key directly in stdout JSON, which can be captured by logs, calling agents, shell history integrations, or downstream tooling. In a skill ecosystem, stdout is often treated as ordinary task output, so exposing long-lived credentials there creates a realistic credential leakage path.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The order flow writes QR-code images into a session directory on disk without clear disclosure, which creates persistent payment artifacts and filesystem side effects that users or calling agents may not expect. In shared or multi-tenant environments, those files could be discovered by other processes or users, revealing payment links or transaction context.

Ssd 3

Medium
Confidence
95% confidence
Finding
The skill requires writing the full API response to a session-based directory under the working directory for every call, even when only a subset of fields is needed. Persistent storage of full responses increases the risk of retaining sensitive business data, identifiers, or user-provided content longer than necessary, and session-based paths can make correlation and later access easier.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.