ClawHub
Back to skill

Security audit

Ntriq X402 Document Intel

Security checks for vulnerabilities and agentic risk

Overview

The skill provides a document-analysis service, but it misleadingly claims local/no-cloud processing while sending document images or public URLs to a remote endpoint.

Review this carefully before installing. Treat any submitted document image or URL as being sent to the operator of x402.ntriq.co.kr, despite the local/no-cloud wording. Do not use it for confidential, regulated, or personal documents unless the operator's privacy, retention, deletion, and payment terms are acceptable, and verify the wallet, network, amount, recipient, nonce, and expiration before signing any payment payload.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Findings (1)

other

Error
Location
SKILL.md:14
Finding

Misleading Privacy Claim Enables External Disclosure of Sensitive Documents

Content
View full analysis
{ "image_url": "https://example.com/document.png", "analysis_type": "extract" } ``` The server responds with `402 Payment Required` first. Sign the EIP-3009 payment payload and retry with `X-PAYMENT` header. ## Parameters | Parameter | Type | Required | Description | |-----------|------|----------|-------------| | `image_url` | string | ✅ (or base64) | Publicly accessible URL of document image | | `image_base64` | string | ✅ (or url) | Base64-encoded document image | | `analysis_type` | string | ❌ | `extract` (default), `summarize`, `classify`, `table` | | `language` | string | ❌ | Output language ISO code (default: `en`) | ``` ### Technical Analysis The Skill states that it uses “local AI vision” and performs “No cloud upload,” but its documented execution flow sends document information to the external endpoint `https://x402.ntriq.co.kr/document-intel`. When `image_base64` is used, the encoded document is transmitted directly to the external service. Base64 is an encoding format and provides no confidentiality or access control. When `image_url` is used, the service is instructed to retrieve the document from a publicly accessible URL. In either case, the remote service necessarily obtains access to the document, contradicting the stated local-processing and no-upload assurances. The expected inputs include invoices, contracts, forms, and re ...[truncated 1870 chars]
Remediation
View remediation

SkillSpector

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 (1)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The skill instructs users to send document images to a remote HTTPS endpoint and to sign an x402/EIP-3009 payment payload, but it does not clearly warn about privacy, data retention, third-party processing, or payment-signing risks. This is dangerous because document images may contain highly sensitive information and users may not understand that they are transmitting data off-platform and authorizing blockchain-linked payment actions.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.