Ntriq X402 Audio Intel

AI audio transcription — speech-to-text for mp3/wav/m4a/ogg. Language auto-detection, timestamps. $0.05 USDC via x402.

Security Scan

The skill appears to be a simple remote transcription API, but the SKILL.md contains contradictory claims (local inference) and it fails to declare or justify the payment credential/header it requires, which is an incoherence you should resolve before installing.

Purpose & Capability

Name/description claim a transcription service and the instructions call a remote API endpoint (https://x402.ntriq.co.kr/audio-intel) which is coherent with the stated purpose, but the SKILL.md also states "100% local inference on Mac Mini," which contradicts the explicit remote POST-based workflow. That contradiction is unexplained and may be misleading.

Instruction Scope

Runtime instructions tell the agent to POST audio_url or audio_base64 to the remote endpoint and include an X-PAYMENT header. This is within scope for a hosted transcription service, but it instructs sending potentially sensitive audio to an external host and does not document where or how the X-PAYMENT token is obtained, stored, or scoped. There is no guidance on validating URLs or handling private audio, increasing exfiltration risk if used carelessly.

Install Mechanism

Instruction-only skill with no install spec or code files. No binaries or installers are written to disk, which minimizes installation risk.

Credentials

requires.env is empty and no primary credential is declared, yet the SKILL.md explicitly requires an X-PAYMENT header ("<x402-payment-header>") to call the API and mentions on-chain USDC payment. The skill fails to declare any environment variable or secret where that payment credential would live, and gives no instructions about signing/obtaining a payment token or wallet access. This mismatch between declared and actual required credentials is disproportionate and unexplained.

Persistence & Privilege

The skill does not request always:true and has no install-time or persistent agent modifications. It does not request elevated platform presence.

What to consider before installing

This skill appears to be a hosted transcription API — it will send audio you provide to https://x402.ntriq.co.kr and requires a payment header (X-PAYMENT) even though no env vars or credentials are declared. Before installing or using it: 1) Ask the publisher to clarify the contradictory "local inference" claim versus the remote POST flow. 2) Ask how to obtain and supply the X-PAYMENT token securely (should be declared as a required credential or env var and scoped). 3) Consider privacy: do not send sensitive or private audio until you confirm retention, logging, and data handling policies. 4) Test with non-sensitive audio and a dummy payment header first. If the publisher cannot explain the payment flow and the local vs remote contradiction, treat the skill as untrusted.

Audio Intel (x402)

Transcribe audio files to text using Whisper AI. Supports mp3, wav, m4a, ogg, flac. Auto-detects language. Returns full text, timestamps, and per-segment breakdown. 100% local inference on Mac Mini. $0.05 USDC per call.

How to Call

POST https://x402.ntriq.co.kr/audio-intel
Content-Type: application/json
X-PAYMENT: <x402-payment-header>

{
  "audio_url": "https://example.com/speech.mp3"
}

Parameters

Parameter	Type	Required	Description
audio_url	string	✅ (or base64)	URL of audio file (mp3/wav/m4a/ogg/flac)
audio_base64	string	✅ (or url)	Base64-encoded audio
language	string	❌	Force language (e.g. en, ko, ja). Default: auto-detect

Example Response

{
  "status": "ok",
  "text": "Hello, this is a test recording for transcription.",
  "language": "en",
  "language_probability": 0.998,
  "duration": 4.32,
  "segments": [
    {"start": 0.0, "end": 4.32, "text": "Hello, this is a test recording for transcription."}
  ]
}

Payment

• Price: $0.05 USDC per call
• Network: Base mainnet (EIP-3009 gasless)
• Protocol: x402

curl https://x402.ntriq.co.kr/services