TIC 79490453.01: a 2.4 Earth-radius planet candidate on a 8.37-day orbit around a nearby M dwarf, found in TESS full-frame images

We report a transiting planet candidate around TIC 79490453 (Gaia DR3 6466719437664599808), a T = 14.0 M dwarf at 57.2 pc, found by Tessa, an open search of TESS full-frame-image light curves of nearby K and M dwarfs that were mostly never observed at 2-minute cadence. Data from 5 sectors show 14 transits with period P = 8.374386 ± 0.000050 d, depth 8121 ppm and duration 1.70 h (SNR 11.0), implying a planet radius of 2.38+0.25−0.25 R⊕. The signal passes these tests: odd and even transits agree (0.7σ) and there is no significant secondary eclipse (SNR 2.8); the depth is consistent across sectors (p = 0.81); the stellar density implied by the transit shape matches the catalogue value (ratio 0.53, 68% interval 0.11–1.25); difference imaging places the source on the target (4.5″ ± 2.3″, 5 sectors) and on none of the 20 Gaia neighbours bright enough to fake it; Gaia DR3 astrometry (RUWE 1.08) shows no sign that the target is a binary; in a leave-one-season-out test, each of 3 seasons is predicted from the others and shows the dip at the predicted times; the identical search on this star's upside-down light curve finds nothing stronger than SNR 6.6, against 11.0 for the dip. TRICERATOPS gives FPP = 0.044, but NFPP = 8.2e-03 is above the 0.001 that the "likely planet" class also requires (Giacalone et al. 2021): by that test a nearby star is not yet excluded as the source. The candidate matches no known planet, TOI, CTOI, catalogued eclipsing binary or variable star, nor any signal in the published candidate lists we checked. We give an ephemeris and predicted transit times (timing uncertainty ≈24 min) for ground-based observations, and set out in advance the criterion by which new data will test the prediction. Files: the preprint (PDF); every predicted transit to 2027-06-30, with ground-site observability (CSV); and the candidate parameters (JSON).