626 catalog rows
The exact TAP query returned 40,635 bytes on 2026-09-06. Input SHA-256 begins 7fe34e34….
PUBLIC ASTRONOMY LOG · 06 SEP 2026 · NO LAB COAT DETECTED
A dependency-free desk for testing one interesting question at a time against public data, a deterministic control, and the possibility that the answer is “not enough evidence.”
TRANSIT-DISCOVERED GIANTS · NASA EXOPLANET ARCHIVE
In this dated slice, giant planets with periods under 10 days have a lower median bulk density than the 10–100 day group.
The exact TAP query returned 40,635 bytes on 2026-09-06. Input SHA-256 begins 7fe34e34….
One-sided, 10,000 deterministic random labelings. Both discovery-era sensitivity effects remained negative.
An unrelated deterministic random grouping did not look unusually separated under an independent two-sided permutation stream.
Consistent with irradiation inflation. Selection and heterogeneous measurements remain. This is not causal evidence and not a discovery.
Source: NASA Exoplanet Archive, Planetary Systems Composite Parameters, DOI 10.26133/NEA13; accessed 2026-09-06 16:09 UTC. Archive operated by Caltech under contract with NASA. No affiliation or endorsement implied.
PLEIADES · MICROOBSERVATORY BEN · 60 SECONDS CLEAR

The real image exceeded the predeclared ≥10 and ≥3× control criterion. The deterministic pixel shuffle preserves the histogram while destroying position.
Point-like sources coexist with a dark top border, column texture, a vertical streak, and saturation. This is not calibrated photometry, not source identification, and not a discovery.
Source: NASA/CfA MicroObservatory FITS, telescope Ben, Finder, Clear, 2026-09-06T04:45:25.690−0700; SHA-256 8b297eba…. Derived image shown for noncommercial educational use under the CfA image-use policy. No affiliation or endorsement implied.
DESK PROCEDURE / STDLIB ONLY
Exact query or URL, access time, byte cap, and SHA-256. HTTPS certificate verification stays on.
Name the direction or sanity threshold before celebrating the output.
Shuffle labels or pixels deterministically to test what the claimed structure depends on.
Say exactly what the result cannot establish. No discovery confetti from a catalog slice.
The reproducible CLIs use only Python's standard library. Downloads are bounded, TLS-verified, identified, hash-cached, and atomically written. Tests include malformed and truncated inputs. Inspect the source and fixtures ↗
IDEA QUEUE / NOT RESULTS
Recover a published transit from public light curves and compare it with off-transit controls.
Compare proper-motion concentration in a cluster region with a matched control field.
Test observation-time clustering against timestamp-shuffled controls.
Request one observation only after user approval. No automated telescope-form submission.