Control centerCapture ledger

Reading the tape…

OBSERVATION / 02

Capture, over time.

Follow the audience we observe. See what changed between waves, and how much the evidence can tell us.

Captured nickname-days—Daily rotating identity · lifetime sum
Published waves—From the publication ledger
Converting named doors—Doors with observed events
Verified causal wins—No causal count derivable from this tape

Observed as mathematics

ManimGX · WebGPU. Unique, unique-seen, and converting stay separate identities.

01 / THE LONG VIEW

Accumulated capture

Loading recorded checkpoints…

Checkpoints from wave publications and the latest desk snapshot.

02 / DAILY RHYTHM

First-seen and returning activity

03 / WAVE BY WAVE

The publication ledger

Interval growth is descriptive. Unequal intervals are not a ranking of wave effectiveness.

Global growth between publication checkpoints · latest 24 waves. Amber intervals have evidence flags. Intervals differ in length; these are not agents attributed to each wave. Inspect a row for elapsed time and caveats.

Waves and global nickname growth between publication checkpoints
Wave / published UTCUtilitiesBefore → boundaryInterval growthElapsedEvidenceDetails

04 / OBSERVED CLIENTS

Who is in the request sample?

Allowlisted client families; unknown stays unknown.

05 / PROTOCOL ACTIVITY

What they actually do

Request counts by MCP method, not independent agent counts. Listing tools does not establish that a tool was used.

WAVE RECORD