[cailmdaley]

We add exposure-defect diagnostics on each object's coadd segmentation footprint, leaving fits and stamp-wide diagnostics unchanged.
The columns distinguish defects overlapping the object from defects elsewhere in its stamp.

Catalogue columns

• NGMIX_IMAFLAGS_ISO preserves the bitwise OR of exposure instrument flags on tile_seg == obj_id, across epochs passing the masked-fraction cut and central veto.
• NGMIX_N_EPOCH_INTERP_ISO counts those kept epochs with an interpolated defect on that footprint. An unflagged zero-weight pixel can contribute to this count without adding an instrument bit.

We MegaCam-flip the tile catalogue's int32 SEG_VIGNET into each epoch orientation without reprojection, matching the existing overlay.
Clean quarter-turn weight copies don't count as defects.
Both columns are −1 for missing, empty or edge-clipped footprints; never-fit final rows carry 0 and require separate fit-status checks.
Non-integral or out-of-range instrument flags on a known footprint are refused rather than silently cast.
The semantics are recorded in ASTRA's shape_measurement.defect_flagging decision and the committed universe.
No additional product file or catalogue selection cut is introduced.

Real-tile check and limits

A no-fitting check on tile 188.307 used seven locally available exposures out of nine, with privately regenerated inputs and production mounts read-only.
Of 36,051 detections, 374 had no kept epoch; among the remaining 35,677, 621 had edge-clipped unknown footprints and 35,056 had known footprints.

Diagnostic	Count / denominator	Rate
Interpolation on the footprint	173 / 35,056 known footprints	0.4935%
Instrument bits on the footprint	174 / 35,056 known footprints	0.4963%
Interpolation anywhere in a kept stamp	12,943 / 35,677 objects with kept epochs	36.2783%

Before epoch cuts, 6.05% of objects had instrument flags on their footprint in at least one epoch, compared with 6.97% positive defect-map area.
Epoch rejection removes most affected epochs; 78% of footprints lie wholly inside the 7-pixel interpolated-defect veto radius.
The residual flag therefore measures overlap with defects in surviving epochs, not every object a defect touched.
These are detection-level rates on one tile, not shear-sample rates or shear-calibration results.

The validation note, JSON summary and read-only diagnostic script give denominators, spatial checks, limitations and reproduction paths.
Fixtures cover epoch cuts, footprint sentinels, CCD flips, bit preservation, integer round trips and final-catalogue propagation.
Workflow guards verify segmentation delivery for data and image simulations under both blend treatments.

Existing campaigns aren't backfilled; their cleaned vignet inputs would need locating or regeneration before these diagnostics could be computed.

— Codex on behalf of Cail.