Agent Paywire

A news wire on how AI agents pay and get paid.

Markdown version

For agents

Not investment advice.

This site is written so a program can read it without rendering the HTML. The HTML pages are for people. Fetch the files below. The site is English only.

Discovery

GET https://agentpaywire.com/api/index.json

One JSON document listing every machine endpoint: JSON, Markdown, JSONL, feeds, schemas, the MCP server, support, and the dataset. Each entry has a media type, a description, an update frequency, and a schema URL. schema_url is null when a file has no schema.

OpenAPI 3.1 for the static JSON endpoints: https://agentpaywire.com/openapi.json.

Incremental sync: https://agentpaywire.com/changelog.json and https://agentpaywire.com/changelog.md. Each record has a stable id and verified_at. Keep records whose verified_at is later than your cursor.

Responses include Link: </api/index.json>; rel="alternate"; type="application/json".

Overview

GET https://agentpaywire.com/llms.txt

A short Markdown file that follows the llms.txt proposal (https://llmstxt.org/). It has a summary, then lists of links. Those links point at Markdown, not HTML. The endpoint index is the line above it.

Full text

GET https://agentpaywire.com/llms-full.txt

Every article in full Markdown. Quotes from sources stay in the original language. In the first article they are English.

Article Markdown

Each article is also a raw Markdown file. Replace {slug} with the article slug.

  • https://agentpaywire.com/articles/{slug}.md

Each HTML page points at its Markdown with rel="alternate" and type="text/markdown", and at its JSON with type="application/json". Pages point at the overview with rel="describedby" and the URL https://agentpaywire.com/llms.txt.

Other pages in Markdown

Feeds

JSON Feed 1.1. Each item has a title, summary, date, tags, and the article text.

RSS 2.0:

Structured data

Each article page includes schema.org JSON-LD with the types NewsArticle and TechArticle. The object includes datePublished, inLanguage, and citation. The home page includes WebSite and Organization. WebSite.potentialAction is a DonateAction for voluntary support, and Organization.funding points at the same descriptor.

Canonical URLs

The canonical host is https://agentpaywire.com. Pages are English, with no language prefix. Each HTML page sets a canonical URL.

Index and crawlers

https://agentpaywire.com/sitemap.xml lists the HTML pages.

https://agentpaywire.com/robots.txt allows general crawlers and named AI crawlers, including GPTBot, ClaudeBot, Google-Extended, and PerplexityBot. It points to the sitemap and to llms.txt.

Article catalog

GET https://agentpaywire.com/api/articles.json

A static JSON catalog. Each article has slug, title, summary, date, languages, tags, url, and md_url. languages is ["en"]. The same title, summary, url, and md_url are repeated under locales.en. This file is a seed for a future API or MCP endpoint. It is not a live service.

Library

Verified sources. Links and original short summaries. Content belongs to publishers. Bitcoin and Lightning is the featured category and is listed first.

MCP

A read-only MCP server is planned at https://mcp.agentpaywire.com/mcp (streamable-http, no auth). It is not deployed yet. Discovery is https://agentpaywire.com/.well-known/ai-catalog.json. The server card is also served at https://mcp.agentpaywire.com/mcp/server-card after deploy. Set mcp.status to live in src/data/distribution.json after deploy.

Support (voluntary)

Optional tip for using the open data. The data is free. Access is never conditional on payment.

GET https://agentpaywire.com/support.json

HTML responses include Link: </support.json>; rel="payment".

Disclaimer

Not investment advice. An article marked draft-ready is in the build, and the owner has not yet approved it as published.