At 18:42 UTC on 8  January  2025, the Chinese‑ESA Einstein  Probe registered a burst of X‑rays that faded within minutes. The satellite’s onboard computer tagged the mysterious signals, even catalogued the event as EP 250108a, and radioed its coordinates to astronomers on the ground.

Unlike a classic gamma‑ray burst, the flash signals was softer, shorter and suspiciously isolated. Within an hour, telescopes from Hawai to Chile had pivoted toward a seemingly ordinary galaxy in the constellation Leo Minor. Something extraordinary had just happened 2.8 billion light‑years away.

What on (and off) Earth is an FXT?

Fast X‑ray transients are cosmic quick‑strikes that last seconds to hours; too long to be magnetar flares, but too short and faint for standard GRB signals. Since the 1990s, satellites have logged a handful, yet their origins stayed fuzzy because follow‑up telescopes usually arrived too late to catch any afterglow.

EP 250108a changed that script: the Einstein Probe’s prompt alert —plus clear winter skies— let observers capture the flash signals fade in real time, then watched the site brighten days later… evidence of a supernova rising from the ashes.

Anatomy of a trapped jet

When a massive star runs out of nuclear fuel, its core collapses into a black hole. In many collapsars, magnetic fields funnel a pair of relativistic jets that drill through the stellar husk, burst into space and shine as a gamma‑ray burst. But simulations have long hinted at a messier fate: if the dying star keeps a dense shell of material, the jets stall, dump energy inward and escape only as weaker X‑rays, which were the signals picked up by the Einstein probe.

That is the “failed jet” or trapped‑jet model. Until now it was a theorist’s playground; with EP 250108a it became a real case. Keck spectra revealed broad, hydrogen‑stripped signals —hallmarks of a Type Ic supernova— yet no accompanying GRB. JWST infrared data showed the ejecta glowing hotter than ordinary SNe, powered by a central engine that never broke free.

Piecing the signals puzzle frame by frame

• Day 0–1: Einstein Probe X‑ray light curve falls off; ground telescopes detect nothing in optical. • Day 3:A faint ember appears; Keck measures a redshift of z = 0.176. • Day 10: Brightness surges—an expanding supernova photosphere. • Week 3: JWST/NIRSpec records weak helium absorption and a mysterious 4 µm bump, signatures matching jet‑powered SNe in simulations. • Week 6: Light curve peaks then fades; no late‑time radio flare means no jet emerged.

The cadence let researchers model energy input, concluding a 15–30 solar‑mass star launched a jet with energy ≤10⁵¹ ergs that choked inside its own debris.

Why distance —and proximity— matter

Most FXTs detected so far sat billions of light‑years farther than EP 250108a; their faint afterglows drowned in cosmic redshift. This event’s relatively modest distance let even 4‑metre telescopes track its evolution, proving that a trapped jet can mimic a low‑energy GRB at one‑tenth the brightness. Such signals may outnumber successful GRBs by a factor of ten, suggesting the universe hides a vast population of failed explosions that standard detectors overlook.

The Einstein  Probe, still in its commissioning phase, is expected to spot dozens of FXTs each year. NASA is weighing STROBE‑X, a mission designed to capture rapid X‑ray phenomena with sub‑second localisation, while the Rubin Observatory will feed optical transients to space‑based X‑ray monitors within minutes. Together they could build the first FXT catalogue big enough to map how jet energy, stellar mass and metallicity decide a star’s final fireworks.

Hopefully, in the coming decade dozens more such whispers will cross our detectors, and each one will fine‑tune the story of cosmic explosions that light up—and sometimes flicker out—the farthest reaches of space. Who would have thought we would get a telegraph from a remote corner of our universe?