Agent Commerce Migration Guide: Retrofit Your REST APIs for Autonomous Agent Buyers

Security checks across malware telemetry and agentic risk

Overview

This is a non-executable migration guide whose payment, credential, and API examples match its agent-commerce purpose, but the examples should be reviewed before production use.

Treat this as implementation guidance, not ready-to-run production middleware. Use sandbox or least-privilege credentials, do not hardcode or commit signing keys, review caller classification so normal clients are not forced into payment flows, and require explicit approval before enabling live payment verification or settlement.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Description-Behavior Mismatch

Medium
Confidence
92% confidence
Finding
The skill is labeled as non-executable educational material, but it embeds detailed operational code that performs network requests, credential use, payment verification, auth bridging, file writes, and proxy configuration. This mismatch can cause users or downstream systems to under-classify the risk and treat the content as harmless documentation when it actually contains copy-pastable production-grade integration logic.

Intent-Code Divergence

Medium
Confidence
86% confidence
Finding
The guide says users can get started with no API key, yet most code paths assume an authenticated GreenHelix client and send Authorization headers using a configured API key. This inconsistency can mislead users into inserting real credentials prematurely or misunderstanding which examples are safe to run anonymously versus which require privileged access.

Missing User Warnings

Low
Confidence
80% confidence
Finding
The manifest advertises required credentials, including an API key and an agent signing key, but does not pair that with strong guidance on secret storage, non-disclosure, and avoiding production credentials in examples or test environments. In a migration guide centered on payments and identity, casual handling of these secrets raises the chance of key leakage, misuse, or accidental inclusion in logs and config files.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal