Echolot API v1

Our own derived data: 12-language press intelligence, trends per country, food and horoscope — the same data as our MCP server and RSS feeds. Machine-readable schema: /v1/openapi.json (OpenAPI 3.1).

Authentication

Send your key in the X-API-Key header (or Authorization: Bearer …). Create keys at echolotnews.com/account; the same key works for MCP. Without a key you get a demo quota of 100 requests/day per IP and 10/minute.

curl -H "X-API-Key: eck_…" "https://api.echolotnews.com/v1/trends/google?geo=HU"

Browser use leaks your key. CORS is open, but anything you ship to a browser is public — call the API from your server.

Responses

{"ok": true, "data": …, "served_at": "…", "generated": bool, "api": "v1"} — data has the same shape as the MCP tool. Errors: {"ok": false, "error": {"code", "message"}} with 400 / 401 / 404 / 429 / 503. 429 and 503 carry Retry-After. Use If-None-Match with the ETag to get 304 when nothing changed — please poll politely.

generated: true marks machine-written prose (summaries, horoscope texts). Recipes under CC BY-SA carry source, source_url and license: attribution is required. News endpoints return our summaries, metadata and links only. Horoscopes are entertainment, not advice.

horoscope

Daily, weekly and monthly horoscopes grounded in the computed sky (JPL DE421). Entertainment, not advice.

• GET /v1/horoscope/day — Daily horoscope + the computed sky of the day
  date, sign, lang RSS: /rss/horoszkop?idoszak=nap
• GET /v1/horoscope/week — Weekly horoscope + the computed sky events of the ISO week
  week, sign, lang RSS: /rss/horoszkop?idoszak=het
• GET /v1/horoscope/month — Monthly horoscope + the computed sky events of the month
  month, sign, lang RSS: /rss/horoszkop?idoszak=honap
• GET /v1/sky — The raw computed sky (JPL DE421): positions, retrogrades, Moon, aspects
  date

food

Menus and recipes in 12 editions; EU-14 allergens (LLM ∪ deterministic dictionary, with evidence). CC BY-SA recipes carry source + licence — attribute them.

• GET /v1/food/menu — The day's menu (allergen-safe: LLM labels ∪ deterministic dictionary)
  date, diet, allergens, cuisines, meals, insulin_resistance, lang RSS: /rss/receptek
• GET /v1/food/week — 7-day menu (weekdays simpler, public holidays special)
  week, diet, allergens, cuisines, meals, insulin_resistance, lang
• GET /v1/food/shopping-list — Shopping list for a week's menu, by shop department
  week, diet, allergens, cuisines, meals, insulin_resistance, lang
• GET /v1/food/recipes/{id} — One recipe: ingredients, steps, allergens + evidence, source and CC BY-SA licence (attribution required)
  id*, lang
• GET /v1/food/recipes — Search recipes by title/description
  q*, lang

trends

Trending signals — our own derived data per country: press corpus velocity, social engines, Google Trends, Wikipedia, Google News, trend matrix, attention gaps.

• GET /v1/trends/corpus — What the PRESS is accelerating on (our 12-language corpus): entities with velocity
  days, limit, language, sphere
• GET /v1/trends/social — Social trending engines: Reddit, X or YouTube transcripts
  source*, language, days, limit
• GET /v1/trends/google — Google Trends 'trending now' search signal per country
  geo, hours, limit, day
• GET /v1/trends/wikipedia — Wikipedia pageview top list per language edition
  geo, limit, day
• GET /v1/trends/news — Google News top stories per country
  geo, limit, day
• GET /v1/trends/matrix — Trend matrix per country: Google Trends × Wikipedia × Google News × our press corpus — consensus / blind spot / echo cells
  geo, day
• GET /v1/trends/attention-gap — Attention gaps: what people search for that the press barely covers, per country
  geo*, lang, limit, min_gap

signals

Business questions answered from our own derived data: how much is X discussed and where, what goes together with X, and what the audience wants that the press does not cover. Press, X (media/political accounts), Reddit and YouTube transcripts, matched to the same entity everywhere.

• GET /v1/signals/mentions — How much is X talked about — per corpus: press, X (elite discourse), Reddit, YouTube transcripts; same entity matching everywhere
  q*, days, lang, niche
• GET /v1/signals/cooccurrence — What goes together with X — words over-represented in the documents about X, in the press, on X, on Reddit or in YouTube transcripts (lift vs the same corpus and window)
  q*, source, days, lang, niche, limit
• GET /v1/signals/demand — Where and how much does the audience care about X: Wikipedia pageviews, Google Trends, Reddit and YouTube mentions — against press coverage
  q*, geo, lang, days
• GET /v1/signals/gap — Attention gaps per country: what the audience searches, discusses on Reddit or watches on YouTube, but the press barely covers
  geo*, lang, limit, min_gap, days

news

Stories, sections, the daily edition, search, sources, economy. Our own output only: summaries, metadata and links — never third-party article bodies.

• GET /v1/stories — Top story clusters of an edition (our summaries + metadata + links; QID, sphere and framing fields)
  lang, sphere, limit RSS: /rss
• GET /v1/stories/{id} — One story: summary, sources, bias distribution, languages, international coverage (metadata + links only)
  id*, lang RSS: /rss/sztori/{id}
• GET /v1/sections/{section} — A section feed of an edition (tech, economy, sport, …)
  section*, lang, limit RSS: /rss/rovat/{section}
• GET /v1/edition/daily — The daily edition (Lapszám) — the day's brief
  lang, limit RSS: /rss/nap
• GET /v1/search — Search the 12-language press corpus (metadata + links; no article bodies)
  q*, days, language, sphere, limit RSS: /rss/kereses
• GET /v1/sources — The source catalogue (written press)
  lang, category, limit, offset RSS: /rss/katalogus
• GET /v1/economy — Economy: the day's macro block of an edition (stored) + the economy section's stories
  lang, date

entities

People, organisations and places (Wikidata QIDs) in the 12-language press.

• GET /v1/entities/top — Top named entities in the news flow with aggregate sentiment
  days, type, language, limit
• GET /v1/entities/{qid} — How a person/org/place is covered: spheres, sources, sentiment, narrative role
  qid*, days
• GET /v1/entities/{qid}/trend — Mention time series + trend signal for one entity
  qid*, window
• GET /v1/entities/{qid}/stories — Stories about one entity across all 12 editions
  qid*, lang, limit RSS: /rss/entitas/{qid}

context

Context signals (weather).

• GET /v1/weather — Weather context for a point + fire-danger index (Open-Meteo)
  lat, lon, place, days, lang