# 490d C422 — Source and component audit of the new 2K profile

24 September 2026. Step 9 of 10 in the C414–C423 authorization.

**All23 relevant source records reconstruct exactly from current source passages.** Twelve excerpt cards retain the actual text and hashes for person baselines, Flood tables, variants, SP locality and qualifications, LXX removal, both500/502 modes, E/P/E² scalars, J's gap-only status and the January Mirror/component convention.

The reconstructed set contains the eight main person stations, all twelve event records in C418's three-Gear boundary controls, the new MT Flood-close2649 endpoint, and the SP-primary/LXX-native Noah rail controls at3900. Person and event calculations use their own independently parsed baselines; no Flood close is typed as a person companion. LXX already contains native Cainan; theoretical OFF uses subtraction rather than a second insertion.

The source passages explicitly warrant endpoints, baseline/variant rules, qualified950/600 lifespans and500/502 bindings. The particular1058 cross-context profile, its finite uniqueness, the mixed-boundary five-station comparison and the held3899 squared-E endpoint are derived arithmetic. No prescription for that combined selection is established by the reviewed excerpts. SP primary Noah death retains its inclusive/harmonic qualification; the new main profile uses SP companion.

Inherited quarter-offset components provide a separate control. Across five station pairs and four upper/lower assignments each, exactly ten comparisons pass: the two matched-offset components per station. This follows from the translation's slope1. Opposite offsets differ by half a year. These are existing signed-coordinate component diagnostics, not newly assigned civil dates or phases.

For the fixed held3899 E² map, all four input/output quarter-offset assignments fail. The two matched offsets have errors−24/529 and+24/529; the opposite assignments give−577/1058 and+577/1058. Jointly shifting both pivot and input by the same quarter offset does return the correspondingly shifted output, but it changes the pivot. Those two comparisons are not one fixed-pivot physical-time action.

The source audit therefore supports the finite person-profile translation and exact scalar endpoint while preserving their different scope. It adds no source point, phase, global Mirror axis or canonical operation.

## Verification and limits

64 main checks and 21 frozen-input/predecessor authentication checks passed. A second main replay produced byte-identical DATA/RESULTS. These checks and the source interpretation were performed by the root agent; no new independent-agent review is claimed. All frozen originals and standing controls remain unchanged. Verification counts are not independent-witness counts.

The alternate component comparisons are not new witnesses, phases or calendar prescriptions. No independent-agent review is claimed. Regular Gears remain Noah/Shem/Flood only.

Mirror/529 focus: K=529, E=25/23, P=70/69. Global paired Mirror January AD1; local junction January AD2; no civil year zero. Regular Gears are Noah/Shem/Flood only. SP minus-one person companions remain local to SP Noah/Shem; Flood close is a separate boundary. No upstream/cumulative Gear, new endpoint or phase, fitted anchor, global axis, primer reopening, canonical or graph edit. Capture now; canonize later. Source identity, context, modal binding, generated coordinates and phase ancestry stay distinct.

## Decision after this result

**C423, proposed and unexecuted:** Consolidate C414–C422, run a separate root-agent arithmetic/source implementation and sealed-packet replay, then publish the C423 synthesis, final checkpoint and recovery packages. Leave C424 unexecuted.

The new profile, harmonic endpoint, source-path obstruction and component scope are now bounded. The tenth authorized step should preserve and verify the completed chain rather than open another research branch.