How Arkitekt works
Arkitekt is an evidence-based structuring tool. You bring real events. It structures them into something that can be proved, disproved, and accounted for.
Four steps: enter the scenario from documents, messages, or video; fill the gaps by probability; select the single variable being tested; then prove or disprove the hypothesis. The reasoning behind the tool is on the Fully Litt A.I. page.
It is not complete. It is being built in the open — and you can help build it.
What you bring in
Documented conversations
Messages, transcripts, correspondence — what was actually said.
Video and recordings
Footage of the event itself, as captured, not as described.
Events and their impact
What happened and who it landed on — from any timeline, recent or historical.
The four steps
- 01Enter the scenario
You load the real events — documents, media, timeline. Nothing is assumed yet.
- 02Fill the blanks
Where the record has gaps, the system fills them by probability and marks them as estimated. Estimates never pass as fact.
- 03Select the variable
You choose the one thing being tested — the claim, the actor, the decision under question.
- 04Prove or disprove
The hypothesis stands against the evidence, or it does not. Either result is an answer.
Try it: a worked example
The public is given information in a way that triggers some form of emotional response followed by a order presented as a request to make a decision with in a time frame.
.
Evidence that held
- CONFIRMED Emotional trigger present — the information is delivered to produce a feeling before analysis.
- CONFIRMED Order presented as request — a directive is wrapped in language that appears to invite consent.
- CONFIRMED Time frame imposed — the decision is bounded by a deadline.
Gaps, filled by probability
- ESTIMATED Neutral context missing — no parallel information that would reduce the emotional load. Read as absence of counterbalance, not proof of intent.
- ESTIMATED Source of pressure unknown — who set the frame and why is not documented. Read as external constraint, not confirmed manipulation.
Variable tested
What defines unnecessary harm that removes free will emotionally, mentally, and physically?
Verdict
Three possible readings:
- Confirmed sequence — the pattern matches a known pressure structure.
- Insufficient evidence — the gaps prevent a definitive conclusion.
- Contradicted — if neutral context or voluntary timing appears, the hypothesis weakens.
A fixed example, shown for illustration. Live analysis runs inside Arkitekt.
What you get back
The evidence that held
What survived the test, kept in order and traceable to its source.
The gaps, marked
Every probability-filled blank stays visibly estimated. No invented certainty.
A verdict on the variable
Proved or disproved, with the structure that produced the result.
What cannot be tested cannot be answered for.
What is structured can be checked by anyone.
Let the trash, take out the trash.
Arkitekt
The unnecessary gets named, confirmed, and set aside. What remains is a structure that can be accounted for.
Arkitekt is not finished. Contributions cover subscriptions and credits, or you can co-create the prompting with me. Send a message to help build it.
Evidence first.
Accountability after.