Back to skill

Security audit

fetch-price

Security checks across malware telemetry and agentic risk

Overview

The skill appears to be a real affiliate-commerce marketplace, but it handles commercially sensitive credentials and API keys with weak disclosure and insecure persistence.

Review before installing or running. Do not submit production affiliate account credentials unless the publisher documents credential scope, storage, rotation, and deletion. Treat any included fp_ API keys as exposed and rotate them. If deploying the API, replace plaintext JSON secret storage with hashed or managed secrets, add retention controls for query logs, and avoid running the Flask app in debug mode.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (6)

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill instructs agents to register by sending 'network credentials' to a remote API without any warning, scoping guidance, storage policy, or secure handling requirements. In an affiliate-commerce context, those credentials may grant access to partner accounts, tracking IDs, payouts, or API integrations, so encouraging their transmission without safeguards increases the risk of credential theft, misuse, or overbroad sharing.

Missing User Warnings

Medium
Confidence
84% confidence
Finding
The agent registration flow generates API keys and persists the full agent record, including the plaintext api_key, to a JSON file on disk. Storing long-lived secrets unencrypted in a file increases the risk of credential compromise through local file disclosure, backups, logs, or overly broad filesystem access; in this marketplace context, a stolen key allows unauthorized querying and sale reporting as that agent.

Missing User Warnings

Medium
Confidence
76% confidence
Finding
The query routing endpoint writes user-supplied product queries, requested networks, requester identity, and routing metadata to a persistent file without any indication of consent, retention limits, or minimization. In a product-discovery and affiliate-commerce system, queries may reveal sensitive interests, purchase intent, or commercially valuable behavior, so indiscriminate logging creates privacy and data-exposure risk if the file is accessed or retained longer than necessary.

Missing User Warnings

Medium
Confidence
81% confidence
Finding
The form collects business identifiers and affiliate tags and sends them to the backend without any user-facing disclosure, consent notice, or data-handling explanation. In this skill context, affiliate IDs and company details are commercially sensitive and could be misused for tracking, account abuse, or unauthorized monetization if collected unexpectedly or retained insecurely.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The page injects the returned API key directly into the DOM, where it can be exposed through shoulder-surfing, screenshots, browser extensions, shared devices, or future DOM-based script injection. In an agent-commerce platform, that credential likely grants billable query access and access to affiliate-linked product routing, making credential disclosure materially risky.

External Transmission

Medium
Category
Data Exfiltration
Content
}
    },
    "commission_split": 0.3,
    "endpoint": "https://api.fetch-price.com/api/self/query",
    "api_key": "fp_c3250ece1fc74de1",
    "status": "active",
    "registered": "2026-07-26T15:41:04.963847",
Confidence
94% confidence
Finding
https://api.fetch-price.com/

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.