How it's graded.

Traced to SEC filing cells. Graded in code. Verifiers format.

The number, in code.

Numeric answers are checked in code against a reference value taken from, or computed from, the filed XBRL facts, within a tolerance. The unit and the scale have to match the question. A value outside that tolerance fails.

The verdict, in code.

Core claim checks are scored on the verdict alone. yes or no must match the reference. The explanation is not scored.

The score.

Core is scored entirely by grade.py. Numeric answers are scored within the stated tolerance, and claim checks on the verdict alone. The official score is score = passed/149. grade.py's passed count is that numerator, and grade.py's total is the official score.

What is kept.

A Core task is kept only when every reference fact matches the filing's XBRL instance exactly: exact period dates, dimensions, and unit. Core is the real_v1 verified batch.

Code

MOMENTUS INC.

Current ratio. 10-Q filed 2026-08-11. Task sn_05b9a1732cbd9388. Computed from filed current assets and current liabilities.

Reference value: 13.70
Unit: ratio
Scale: ones
Tolerance: 0.005
Example answer: 13.703

Pass. Example. 13.703 is 0.003 from 13.70, inside the tolerance of 0.005.

Verdict

THUNDER POWER HOLDINGS, INC.

The claim said the current ratio was at least 2. Task cl_06825d0c83d7dc4a. Computed from filed current assets and current liabilities.

Reference ratio: 1.49
Verdict: no

The claim that it was at least 2 is false.

Fail. Example. The reference verdict is no. An answer of yes fails. grade.py scores the verdict alone.

The window.

Filings: Filings are from Q2 2026 or later.
Datasheet: The record of the pack is the datasheet.