DESCRIBEWHAT YOU CANNOT SEE.
A concealed target. A sealed record. A method before interpretation.
SESSION BRIEF.
Before tasking, record anything already present in mind or body. This separates pre-existing noise from session contact.
Record contact, not conclusions. If naming intrudes, declare AOL — ANALYTICAL OVERLAY and continue. Seal the session before disclosure. Compare only after the record is closed.
Nothing entered here is treated as target contact.
TARGET REFERENCE.
The reference is used only to designate the concealed target. It does not describe the target in visible terms. Do not decode it. Do not search it. Record only immediate contact.
FIRST CONTACT.
Write the target reference. Make the first spontaneous mark. Do not refine it. Do not make it look correct. The ideogram is immediate contact, not illustration.
A // FEELING / MOTION
TRACE THE MARKB // MAJOR GESTALT
ONE OR MORE IF NEEDEDIf an object-name arrives, do not follow it. Declare AOL. Continue.
DESCRIBE. DO NOT ANALYZE.
Record raw descriptors before consulting prompts. Do not explain what the target is. If an identity appears, declare AOL BREAK. Naming is preserved separately from signal data.
DRAW. DO NOT NAME.
MOBILITY CUES
PASSIVE FORMORGANIZE WITHOUT CONCLUDING.
Organize the session data without collapsing it into a conclusion. List contact in the proper column. A clean matrix preserves the session. A rushed conclusion degrades it.
REFINE THE LEAD.
Stage V begins from data already recorded in Stages I–IV. Select one earlier signal and interrogate it carefully. Do not introduce a new target identity here. Refinement is allowed; invention is not.
TRANSLATE THE STRUCTURE.
Construct a simple model from contact. Tap a shape to select it, drag to move it, and use the visible controls above to duplicate, delete, or undo. This is not a finished drawing. It is a structural translation of the session.
SEAL THE RECORD.
You are about to close the session before target disclosure. Once sealed, the descriptive record cannot be rewritten. A session fingerprint is generated. The target remains concealed until the record is closed.
THE RECORD DOES NOT MOVE.
Before disclosure, rank four candidate targets from strongest correspondence to weakest. One is the assigned target. The others are decoys. Judge by contact, not preference.
JUDGE BY CONTACT.
Use only the sealed session record. Rank the following still images from strongest correspondence to weakest. Each image depicts a real-world target subject.
Then continue strongest → weakest.0 / 4 RANKED
TARGET DISCLOSED.
—
COMPARE THE RECORD.
A rank-1 result means the assigned target was judged the strongest correspondence among four candidates after the record was sealed. With four candidates, first place has a 25% chance expectation. Single sessions are anecdotal. Patterns across repeated blind trials are more meaningful.
SRI and military programs reported sessions they considered successful, and some experiments used blind matching by independent judges. Later official review concluded the phenomenon was too unreliable and inconsistent for intelligence operations. FIELD 002 tests the method; it does not certify paranormal ability.
This standalone master preloads the target library in the browser, so a technically sophisticated user could inspect source code. It also uses AI-generated ultra-realistic still depictions of real-world targets in place of animated examples; production can swap these assets for vetted real photographs without changing the protocol. The production RDIV build should assign and commit the target server-side, disclose only the encrypted coordinate, and reveal the target only after the session and ranking are cryptographically committed.
A useful session is not always a correct identification. A useful session is a clean descriptive record. This system does not reward retroactive matching. It preserves what was actually said before the answer was known.