Method
Intelligence is not secrets. It is a discipline for turning scattered public facts into a picture someone can act on, without fooling them along the way. Here is the cycle we run, and the rules that keep it honest.
1. Question. Nail down exactly what is being asked, and by whom. "How many public federal terror alerts this year?" is a question. "Is terrorism up?" is a mood. The first is answerable; the second is where people smuggle in their priors.
2. Collection. Gather from open sources only: government pages, official data feeds, published reports, reputable press. No private data, no scraping of databases that forbid it, nothing obtained by pretending to be who we are not.
3. Verification. Prefer primary documents to someone's summary of them. Cross-check anything that matters against a second source. Mark what is confirmed, what is reported but not confirmed, and what is assumed — separately. Never let an assumption wear the clothes of a finding.
4. Analysis. Count, sort, de-duplicate, and look for the gap. The most useful result is often discovering that nobody tallies the thing at all. A gap is a finding, not a failure.
5. Readable output. Write it so a non-specialist can act on it, cite every claim, and state plainly what was and was not searched. The measure of a read is not how impressive it sounds but whether the reader can use it without being misled.
OSINT — open-source intelligence — is what we do: everything from public, lawful sources. HUMINT — human-source intelligence — relies on people and their trust, and carries duties of consent, safety, and law that a public research program cannot take lightly. We do not run human sources here. If a question needs one, the honest answer is that it is out of scope, not that we will try anyway.
A public research program is only worth the name if the public can check it. Everything here is checkable: the links are live, the feeds are real, and the reasoning is on the page. That is the point — not to be trusted, but to be verifiable.