TESS sector 97 excludes a stable 235.587-day repeat with a 0.10-day eclipse core and at least 15% observed loss in EPIC 220208795

We test two proposed repeat periods of the deep K2 eclipse in EPIC 220208795. We use frozen public K2 data, TESS pixel data from six sectors, and ZTF DR24 photometry. The model has a 0.10-day core with at least 15% observed light loss. We allow a reference-time shift of plus or minus 0.02 day and a period rounding range of plus or minus 0.0005 day. TESS sector 97 covers cycle 15 of the 235.587-day repeat. Its 43 nominal-core samples have median relative flux 0.9987. Two complete, separate exposures reject the full period range 235.5865 to 235.5875 days for every allowed reference shift. They pass a five-error brightness rule with a 5% baseline allowance, even when the measured local scatter replaces smaller formal errors. Alternative aperture and background choices support the result. None of 45 added 15% box eclipses is wrongly rejected by this rule. The error rule is not a calibrated false-alarm probability. The 290.230-day range has no usable core exposure in these TESS and ZTF data and remains untested. A saved relative K2 data file has an independently checked 0.05-day origin-label error; we use the original FITS times. The result constrains a stable observed eclipse model. It does not identify a planet or exclude a changed depth, duration, or timing. The package includes source data, code, exact exposure evidence, provenance, figures, and offline reproduction instructions.

EPIC 220208795 · TIC 336889445 · TESS sector 97 · K2 · ZTF DR24 · eclipse recurrence · circumsecondary disk · time-series photometry