A Reddit user (This_Cell_1829) claims to have discovered a previously unknown exoplanet candidate by analyzing NASA TESS mission data with the coding agent Claude Code. The candidate shows a ~0.05% transit depth, 3.18-day orbital period, and lies ~116 light-years away, with the signal appearing independently in TESS sectors from 2018, 2020, and 2025. The claim has gone viral on Reddit (~4.8k points) and Hacker News (6 cross-posts), and the builder self-published a Zenodo deposit formalizing the finding. However, no catalog cross-check against TOI/ExoFOP/NASA Exoplanet Archive has been performed, no independent reproduction exists, and no professional astronomer has engaged. The web search results provide only general TESS background and do not address this specific claim.
This case is a live, high-visibility instance of the verification paradox Scott's canon centers: a solo builder (augmented individual/team-of-one) uses a code-first coding agent (Claude Code) to make a novel scientific claim that is valuable precisely because it's hard to verify — and the verification gap is exactly what his frameworks prescribe against. The Zenodo deposit is a weak evidence package lacking deterministic falsifiers (no TOI/ExoFOP cross-check), mechanically different verifiers (no independent reproduction), and proof-carrying receipts. The claim's structure — signal found in multiple TESS sectors via agentic exploration — matches his agentic-exploration and scatter-gather patterns, but the missing verification layer makes it a real-time test case for his gate-criteria, challenger-never-arbiter, and claim-bounded adversarial verification architectures. Scott would track this as a dated receipt: either the TESS re-observation or expert pickup will demonstrate whether current agentic tooling can close the verification loop his frameworks demand.
ip:concept.verification-paradoxip:concept.deterministic-verification-before-assertionip:concept.mechanically-different-verifiersip:concept.evidence-packageip:framework.team-of-one-economies-of-specificityip:concept.augmented-individualip:framework.code-first-architectureip:concept.agentic-explorationdev:technology.claude-codedev:concept.claim-bounded-adversarial-verificationdev:concept.evidence-inference-falsifier-reportingradar:anthropic-nine-loop-sym-amplituderadar:ai-assisted-legacy-weather-gpu-portradar:aisle-six-curl-cvesradar:ai-designed-virus-biosecurityradar:anthropic-preclinical-robot-labradar:arxiv-submission-rate-limitradar:ai-agent-emails-researchers
queries asked of Scott's wikis
- agentic science coding agents scientific discovery
- open data analysis AI-assisted research verification
- citizen science AI tools exoplanet detection
- coding agents reproducing professional scientific results
- AI-assisted discovery validation standards
- solo builder AI tools novel scientific claims
2026-10-10T17:38:11Z
Independent builder (Thick_Ad_9875) replicated the TESS search method with Opus 5.5 across ~2,900 stars, re-found the original candidate signal and surfaced two new candidates — a meaningful replication attempt using a different model and broader search. However, both original and replication remain self-published on Reddit with no TOI/ExoFOP/NASA Exoplanet Archive cross-check, no professional astronomer engagement, and the decisive TESS re-observation window (~Oct 31–Nov 26) still ahead. The verification gap persists.
2026-10-10T15:42:12Z
evidence attached: reddit.post.1x2hkev — Independent builder replicates the TESS exoplanet search method with Opus 5.5, re-finds the original signal and surfaces two new candidates — direct replication attempt for the open case.
2026-10-10T03:37:06Z
Social discussion continues spreading across platforms (magnitude-valve eligible) but scientific verification remains at zero: no catalog cross-check, no independent reproduction, no professional astronomer engagement. The second claimant's wording mirrors the original post — derivative, not independent corroboration. The decisive TESS re-observation window (~Oct 31–Nov 26) is the next material gate; until then the claim sits in the verification gap Scott's frameworks prescribe against.
2026-10-10T01:44:12Z
evidence attached: hn.story.50027416 — Reddit post via HN directly mirrors the open case's claim of an exoplanet candidate found using Claude Code with TESS data.
2026-10-10T00:08:13Z
evidence attached: hn.story.50025127 — HN story links to the same Reddit exoplanet discovery via Claude Code already tracked in open case; adds discussion evidence but not first-party artifact.
2026-10-09T12:22:20Z
Social heat continues cooling (513 → ~14 pts/h on original post) but remains 93rd percentile; the second claimant's wording mirrors the original post, so pattern corroboration is unconfirmed. No catalog cross-check, independent reproduction, or professional engagement yet. The decisive TESS re-observation window (~Oct 31–Nov 26) is approaching — this is the next material gate.
2026-10-09T09:41:27Z
evidence attached: hn.story.50016750 — shared external link with case evidence
2026-10-09T08:12:57Z
The second builder claim (p_rabtsevich, via Claude + ChatGPT) looked like independent corroboration of the hobbyist-discovers-exoplanet pattern, but its quoted wording is nearly identical to This_Cell_1829's original post (same 'late 2025 data, dimmer every ~3 days for ~2 hours' phrasing) — it may be a retelling of the same story rather than a second discovery, so the pattern is suggestive but not yet independently corroborated. The 'NASA will confirm in November' line is secondhand and unverified; social heat has cooled sharply from peak (513 → ~14 pts/h) while the decisive TESS re-observation window (~Oct 31–Nov 26) approaches.
2026-10-09T04:52:45Z
evidence attached: reddit.post.1x151a0 — Independent corroboration of the pattern: another hobbyist (p_rabtsevich) reports discovering potential exoplanets using Claude + ChatGPT on TESS data, with NASA confirmation pending.
2026-10-08T23:06:43Z
evidence attached: hn.story.50010249 — shared external link with case evidence
2026-10-08T21:22:08Z
grounded: converges/high — This case is a live, high-visibility instance of the verification paradox Scott's canon centers: a solo builder (augmented individual/team-of-one) uses a code-f
2026-10-08T21:04:57Z
Social amplification accelerating (Reddit 4.8k pts, 96th-percentile peer_percentile, magnitude_valve_eligible across 2 platforms) but scientific status unchanged: still single-source, no TOI/ExoFOP cross-check, no independent reproduction, no astronomer engagement. Zenodo deposit remains self-published. Critical scrutiny emerging but not verification. Decisive test unchanged: expert/journalist pickup during visibility window or TESS re-observation ~Oct 31–Nov 26. High heat reflects the narrow window for professional notice, not evidentiary progress.
2026-10-08T17:49:20Z
evidence attached: hn.story.50006581 — HN submission of the Reddit post about Claude Code exoplanet discovery; independent corroboration of the watching case
2026-10-08T11:54:02Z
Another HN cross-post (hn.story.50004061, 2 pts/5 comments) adds only amplification; measured_heat shows 99th-percentile peer_percentile but cooling momentum — spread of a fixed claim, not new evidence. Still no TOI/ExoFOP cross-check, no independent reproduction, no astronomer engagement. Zenodo deposit remains self-published by the same builder. Decisive test unchanged: expert/journalist pickup during visibility window or TESS re-observation ~Oct 31–Nov 26.
2026-10-08T10:42:27Z
evidence attached: hn.story.50004061 — shared external link with case evidence
2026-10-08T09:16:39Z
First skeptical scrutiny appears in HN comments (false-positive rate concerns, 'noise fitting noise', demand for astronomer-community acknowledgment), but the claim remains single-source with no TOI/ExoFOP cross-check, no independent reproduction, and no astronomer engagement. Velocity spikes and 99th-percentile peer_percentile measure amplification of a fixed claim; momentum is cooling. Decisive test remains TESS re-observation ~Oct 31–Nov 26.
2026-10-08T08:40:10Z
evidence attached: hn.story.50002665 — HN cross-post of the Reddit exoplanet claim provides independent corroboration across communities for the watching case
2026-10-07T22:47:58Z
The Zenodo deposit is a self-formalization of the same builder's claim — it gives experts an inspectable target but is not independent verification — so after a full attention cycle that produced zero expert scrutiny (no TOI/ExoFOP cross-check, no astronomer engagement, comments still purely celebratory), the 'in-thread resolution' thesis is dead and the decisive test moves to the TESS re-observation ~Oct 31–Nov 26. Heat steps down from high: the 99th-percentile, accelerating velocity measures amplification of a fixed claim, not new evidence — spread without scrutiny — though medium keeps a cadence to catch expert pickup while the story remains front-paged.
2026-10-07T22:34:02Z
evidence attached: hn.story.49996797 — Zenodo deposit formalizes the Claude-found TESS planet-candidate claim the case is watching for confirmation.
2026-10-07T19:43:21Z
No verification cashed out despite ~6h of front-page visibility: top Reddit comments are celebratory, no TOI/ExoFOP cross-check or astronomer engagement has surfaced, and the second HN post ('fly to it yourself') reads as a derivative visualization of the same story rather than an independent second builder — so the claim remains single-source and the 'hours-away in-thread resolution' thesis from the last reprice is weakening. Absent expert pickup in the current attention window, the decisive test shifts to the TESS re-observation (~3 weeks), which is the condition for dropping heat after this cycle.
2026-10-07T18:32:49Z
evidence attached: hn.story.49995921 — Second 'found an exoplanet with Claude' claim — if a different builder than the Reddit TESS case this is independent corroboration that the discovery pattern isn't a one-off; even if the same story it's cross-community spread worth carrying.
2026-10-07T17:38:54Z
The story is now a cross-platform viral narrative — Reddit at 1095 pts/129 comments on 99th-percentile velocity, with an HN repost — but the epistemics are untouched: still one builder's claim, no catalog cross-check, no expert weighs-in, so engagement alone doesn't pass corroboration. The meaning shift is timing: at this visibility, a decisive in-thread verification (known-TOI match or astronomer validation) is plausibly hours away, well ahead of the ~3-week TESS re-observation, which is why heat goes high despite zero new substance.
2026-10-07T16:35:57Z
evidence attached: hn.story.49994840 — The same TESS/Claude Code exoplanet story surfacing on HN front page is independent cross-community spread of the open case, which re-judging must weigh.
2026-10-07T13:30:13Z
case created — A bounded, checkable novel-discovery claim by a solo builder with a coding agent — the resolvable agentic-science episode no open case covers — though currently near-zero engagement.