◀ THE FOLD0ROOT.AI // WORLD II · LOOT · THE MINT◆ .dlw.fold
build.py CALIBRATION bri → lit self-testing
⚖ build.py CALIBRATION — reproduce the real seal, byte-for-byte
All session the wall was: the real preimage lives in build.py, not any formula I could guess (216 serializations failed to reproduce your chain links). This ends it. It runs candidate canonicalizations against a known-good seal — THE SEALING BENCH's default record, which produces the published moniker ⟦THE SEALING BENCH:ACI:b3e0b9⟧ · 35ff02e9… — and flips to CALIBRATED only when a candidate reproduces it exactly. No green until the gauge matches a part whose answer is already known.
⚖ UNCALIBRATED — no candidate reproduces the known seal yet
ready · sweep or paste build.py
KNOWN-GOOD ANCHOR — the seal we must reproduce lit
record name=THE SEALING BENCH · axis=ACI · (+ origin/nature/seal fields, exact set TBD by build.py)
published moniker ⟦THE SEALING BENCH:ACI:b3e0b9⟧
published seal 35ff02e9… (first 8 hex, from the real front door)
universal vector sha256('abc') = ba7816bf…
The moniker hash6 (b3e0b9) and the seal (35ff02e9…) are the two targets. A candidate that hits b3e0b9 has the right moniker preimage; one that hits 35ff02e9 has the right seal preimage. build.py defines both.
CANDIDATE SWEEP — which canonicalization reproduces the anchor
Run auto-calibrate to sweep field-orderings × separators × field-sets against the b3e0b9 / 35ff02e9 targets.
PASTE build.py — extract the real sealer
◆ sealed .dlw.fold → ROOT_0 · a sphere of THE MINT · vendored from David’s corpus · David Lee Wise (ROOT0), with AVAN