What

Structured JSON records covering AI, semiconductors, and macro. Each record is multi-source verified and scored for credibility.

16 fields: id, title, summary, entities, tags, sentiment, impact_score, credibility_score, methodology, cross_validation_sources, sources, related_ids, language, domain, published_at, expires_at

No raw text. No opinions. Just verified facts your agent can act on.

Why

agent scrapes itself

noisy HTML, no validation, format varies per source, burns your LLM budget parsing junk

signalsfeed

clean JSON, cross-checked facts, honest credibility score, source links for verification

$0.005 per request. Cheaper than one LLM call to parse a raw webpage. No API keys. No signup. No subscription.

How

# 1. Free — see what's new
GET /api/v1/ai?since=2026-04-14T00:00:00Z
→ 200 [ {id, title, tags, published_at}, ... ]

# 2. Pay — get full records (1-20, same price)
GET /api/v1/ai?ids=uuid1,uuid2,uuid3
→ 402 $0.005 USDC via x402
→ 200 [ {all 16 fields}, ... ]

Domains: /api/v1/ai · /api/v1/semiconductor · /api/v1/macro