TIC 116046773: an unconfirmed transit hint at P = 5.6503 d on a K dwarf, from a coherent search of public QLP light curves, with registered predictions for TESS Sectors 117 and later

We report an unconfirmed transit-like signal on the K dwarf TIC 116046773 (Tmag 10.36, Teff about 4210 K, about 62.6 pc) and register predictions for TESS Sectors 117 and later. The signal comes from a coherent multi-sector box-least-squares search of the public QLP light curves of 323 bright K and M dwarfs (sectors up to 96, at least six sectors per star, no TOI or CTOI; the whole stratum, not a sample). It is a dip of 1.5 h at P = 5.650353 d in all seven search sectors (23 transits): 912 +/- 54 ppm in separately reduced TESScut 3x3 pixel photometry (same pixels, same AI family), 1079 +/- 81 ppm in QLP SAP flux. LEO-Vetter 1.2.0 raises no flag in either photometry (MES 12.0 and 14.1) and TRICERATOPS gives FPP 0.033 and 0.059 (NFPP about 1.2e-5; not a validation, no contrast curve). We see no significant second dip once correlated noise is allowed for (largest features 266 ppm at phase 0.26 and 409 ppm at phase 0.67 with the test window masked; empirical z <= 2.9; 3 sigma limit about 431 ppm, 40 percent of the primary depth). The difference-image centroid (3 sectors) is consistent with the target within the realistic error of 2-2.5 arcsec plus 1-2 arcsec of systematics (no statement below about 7 arcsec); it does not exclude a common-proper-motion companion 9.0 arcsec away (Gaia DR3 4608570247290179328, Delta G = 5.9), on which a 12 percent eclipse would produce the dip; adaptive-optics or speckle imaging would resolve it. The hint is not clean: the star FAILED the first hash-registered gate of the search (a harmonic-veto flag that we judged afterwards to be a coincidence), so it is a post-hoc inclusion, picked from 318 searched stars and a ranked list of 7 (in the block-shuffled null 9 of 318 stars reach its coarse SNR of 9.14, against 13 real stars; no significance is corrected for that selection); the QLP depth in Sector 52 is much larger than in the pixel photometry, probably through an over-subtracted QLP background (Sector 53 also has a negative background without such inflation); odd and even transits differ by 2.9 sigma in QLP but by 0.1 sigma in pixel photometry; the LEO-Vetter significance is low; and an eclipsing binary of period 2P or a blend is not excluded. Two other flagged objects of the pilot are an eclipsing binary (TIC 286311568) and a hint that failed vetting (TIC 357170355, reported as a negative result). What is registered: ephemeris v2 (P = 5.650353 +/- 0.000025 d, T0 = BTJD 3451.3131 +/- 0.0045), the table of predicted transit times with 9-11 min (1 sigma) uncertainty for Sectors 117, 118, 124, 125, 131 and 132 (first transit 2027-05-10 16:15 UTC, geocentric, +/- 10 min; the windows of Sectors 124 and later are extrapolated), the statistic (weighted trapezoid depth at the predicted times against 200 foreign phases), the decision rule (SUCCESS, FAILURE or UNDECIDED, including conditions on phase 0.5 and on odd and even transits), the evaluation script (tested on Sectors 79 and 80: SUCCESS; with a shifted phase: FAILURE), the QLP data product, and a short addendum (SHA-256 98dcdccb819e29c040d90da0d0d42074367a65c4be006fad652f705325608701; written on 26 October 2026 before publication and before any sector >= 97 data existed, so it uses no data) that fixes the governing text, the timing of the first evaluation (once, after both Sector 117 and Sector 118 are at MAST), cumulative re-evaluation if UNDECIDED (the first SUCCESS or FAILURE is final), that nothing may be excluded, and that a sector without a QLP light curve is UNDECIDED; it removes the main ambiguities, and three small gaps remain (what happens if Sector 118 never gets a QLP light curve; no fixed calendar day for 'as soon as'; the QLP file version is not fixed), which count as UNDECIDED until a later hashed addendum that can only be written before any such data are loaded. The rule files were hashed with SHA-256 on 26 October 2026 at 10:12 UTC, before any data of a sector >= 97 of this star were loaded (SHA-256 of the ephemeris file: e38804f780f5b49f533a9cf4541d284d0f88793274eb460f3df0587238836f37). The hashes are self-generated; the publication date of this record is the external time stamp, and no data of Sector 117 can exist before 10 May 2027. We claim no priority: ExoFOP lists no TOI and no CTOI for this star on 26 October 2026, and the star is bright enough for QLP-based searches by the TESS teams. The analyses were executed by AI instances (Claude, Anthropic) on public data under the author's direction; the author has reviewed the summary of the results but has not yet independently re-run the analyses, so all conclusions are tentative.