Anything Engine Optimization The rolling record of the AI-search industry.

API & MCP

The wire is built to be pulled. Everything we publish is available as structured data — free, no key, CORS-open. One ask: when you use wire content, cite the item URL. Each item, in turn, cites its original source.

JSON API

  • /api/v1/wire.json — every item: headline, summary, full text, lanes, source attribution, dates, flags, corrections.
  • /api/v1/lanes.json — the lanes with counts and per-lane RSS feeds.
  • /api/v1/item/{id}.json — one item by id.

Regenerated on every publish. Timestamps are ISO 8601 UTC; source_date is the source's own publication date — when we couldn't pin it down, the item is flagged, not guessed.

MCP server

AI agents can connect over the Model Context Protocol (streamable HTTP, no auth):

https://anythingengineoptimization.com/mcp

Tools: wire_latest (limit, lane), wire_search (query), wire_item (id), wire_lanes. In Claude Code:

claude mcp add --transport http aeo-wire https://anythingengineoptimization.com/mcp

Or add it as a custom connector in claude.ai / any MCP-capable client with the same URL.

The record as data

The whole record as JSONL, rebuilt on every publish, CC BY 4.0: /dataset/ — items, figures, claims, docket, glossary, entities, a provenance graph and a change feed (changes.jsonl, filter by date for what changed since). A dataset card and manifest ride alongside.

Agent skill

/skills/aeo-wire/SKILL.md — an installable skill that tells an agent when to consult the record and how to quote it: carry the date, cite the item, preserve the label. Pair it with the MCP server above.

Originated series

  • /crawl-ledger/ — this site's own weekly count of AI-crawler requests against AI-referred visits.
  • /aeo-index/ — who the engines cite for a fixed panel of the beat's own questions, weekly, method open, our own count included.
  • /docket/ — every case and regulatory action on the record, dated.

Feeds

  • /rss.xml — everything, newest first.
  • /lane/{lane}/rss.xml — any lane on its own.
  • /llms.txt — the publication described for language models.
  • /sitemap.xml

Terms, plainly

Use the data, build on it, cite the wire. Don't republish wholesale and call it your own — that's the one thing Anything Engine Optimization exists to never do to anyone else.