Research

Analysis generated across 134,158 indexed articles. Every surface below is served as Markdown, sized to drop straight into an LLM context window — no scraping, no HTML to strip.

Topic briefs

A 30-day picture of one topic: what happened, which entities drove it, which sources covered it. See an example →

Entity dossiers

Everything the corpus holds on one company, model or person: all mentions, co-occurring entities, topic distribution over time. See an example →

Timelines

Events in chronological order, filterable by topic and window — the last 14 days, or one topic over 30.

Comparisons

Side-by-side breakdowns of sources, topics, entities and tooling live at /compare.

Questions

What is a topic brief?
A 30-day picture of a single topic: what happened, which entities drove it, and which sources covered it, assembled from every indexed article matching that topic.
What format are the research pages served in?
Markdown. They are sized to paste straight into an LLM context window, so an agent fetches one document instead of scraping article pages.
How current is the data?
The corpus ingests every 30 minutes, and each brief is generated at request time, so a page reflects the index as of the moment you load it.
Can an agent use these directly?
Yes. Fetch the URL and read the Markdown, or reach the same data through the MCP server and the endpoints listed in llms.txt.