QLP v01 spline flux retains about 6 percent of the Sector 7 dip of TIC 110969638 in fixed-shape fits

We test a reported TESS exocomet candidate with public Sector 7 images and archived light curves. We fit one time shape from ELEANOR PCA flux, then hold it fixed. With a linear baseline over four days, the QLP v01 raw depth is 8934 plus or minus 686 parts per million. Its spline depth is 492 plus or minus 636 parts per million. Shared residual blocks give a raw-minus-spline difference of 8441 plus or minus 690 parts per million. The fitted retained fraction is 0.055 and ranges from 0.035 to 0.074 across four tested time baselines. These numbers compare saved columns; no signal was injected through QLP. A single source with the official TESS pixel response function has a fitted offset of 0.245 pixel from the registered target. Adding derivatives for a common image shift changes the offset to 0.383 pixel. Both models leave more spatial residual than all 2000 draws from the tested shared-block noise model. Four nearby TGLC extractions contain the same dip, but they share image data and permit variable-source leakage. Thus, these fits do not establish a host or a physical cause. The result concerns one event and the saved product versions. It establishes no planet or exocomet. The package preserves 12 input files, source URLs, hashes, code, fit tables, model checks, figures, and reproduction instructions.

TIC 110969638 · TESS Sector 7 · exocomet candidate · QLP v01 · difference imaging · pixel response function · blended sources · reproducible astronomy