# AgentShelf OG Audio Extractor

> AgentShelf OG Audio Extractor is a paid API for AI agents from agentshelf.syntexa.ch, paid per call via x402, $0.01/call, status unknown (last checked 2026-10-02).

Extracts the og:audio Open Graph meta tag value from a public web page URL or raw HTML string, in an SSRF-safe manner.

## Facts

- Endpoint: POST https://agentshelf.syntexa.ch/v1/profile-og-audio?utm_source=zero.xyz
- Price: $0.01/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-02
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/agentshelf-og-audio-extractor-3798fe14
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_4rWdQlPsz6gH-r_vv04ss

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability agentshelf-og-audio-extractor-3798fe14 -d '<json body>'
```

Example prompt: Can you pull the og:audio value from https://example.com/podcast-episode — I need to know what audio URL that page declares in its Open Graph tags.

## When to prefer this

Use this endpoint when you specifically need the og:audio Open Graph property from a public webpage or raw HTML, and want SSRF-safe server-side fetching without building your own HTML parser. Prefer the free sandbox endpoint (/v1/sandbox/profile-og-audio) for testing or non-production use; use this paid endpoint for reliable production extraction. Choose this over general-purpose HTML scrapers when you only need the og:audio field specifically and want a deterministic, low-cost single-purpose extraction.

## Known failure modes

- URL is unreachable or returns a non-200 HTTP status — endpoint may return an error or empty result
- Page has no og:audio meta tag — result will be null or empty
- URL targets a private/internal network address — blocked by SSRF protection
- HTML input exceeds 200,000 character limit — request rejected
- URL exceeds 2,048 character limit — request rejected
- Malformed HTML may cause the parser to fail to locate the meta tag

## How this service works

Call when an agent needs og:audio extracted from a public page or HTML (SSRF-safe). Exact $0.01 USDC. Prefer unpaid POST /v1/sandbox/profile-og-audio first.

## Output

Returns the value of the og:audio Open Graph meta property found in the page's HTML — typically a direct URL to an audio file. If no og:audio tag is present, the response will indicate an empty or null result. The extraction is performed in an SSRF-safe environment, so private or internal network addresses are blocked.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "url": {
   "type": "string",
   "maxLength": 2048,
   "minLength": 8,
   "description": "Public page to fetch. The selector or profile is fixed by the SKU. Provide url or html."
  },
  "html": {
   "type": "string",
   "examples": [
    "<!doctype html><html lang=\"en\"><head><title>Hello</title>\n<meta name=\"description\" content=\"Desc\"><meta property=\"og:title\" content=\"OG\">\n<link rel=\"canonical\" href=\"https://example.com/\"><link rel=\"icon\" href=\"/favicon.ico\">\n</head><body><h1>Hello</h1><a href=\"https://example.com/a\">A</a></body></html>"
   ],
   "maxLength": 200000,
   "minLength": 1,
   "description": "HTML to extract from locally. Provide html or url. When both are set, html is used."
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/agentshelf-og-audio-extractor-3798fe14/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from agentshelf.syntexa.ch](https://www.zero.xyz/host/agentshelf.syntexa.ch/llms.txt)