How We Verify
Methodology states the controls. This page records dated tests of those controls and reports only the results supported by the retained build record.
These are pre-publication findings, not corrections. A claim that reached a subscriber and was wrong belongs in the Corrections Log.
Validation record
| Date | Test | Sample | Seeded defects | Additional defects | Result | Evidence |
|---|---|---|---|---|---|---|
| Independent source check | Five event drafts, followed by a focused check of the 24-page Colorado source | 6 | 3 | All six seeded defects were found across the two runs. | ||
| Registry verification | Arkansas SB 258 status history and final engrossed bill | 0 | 1 | The proposal was recorded as never enacted and excluded from published counts. |
What the source-check test showed
On 19 July 2026, one check covered five event drafts and found four of six seeded defects. A separate focused check read the full 24-page Colorado source for the remaining item and found the other two. Across the two runs, the process found all six seeded defects and three additional issues. The contrast matters: a broad batch found title and summary drift, while the focused source review found deeper qualifier omissions. The operating procedure keeps the independent source check scoped to one item and its full source.
Worked registry example: Arkansas SB 258
A registry verification on 14 July 2026 checked the official status history and final engrossed bill for Arkansas SB 258. The proposal failed third reading twice and died at sine die adjournment on 5 May 2025; its proposed effective date never became operative. The registry records it as never enacted and excludes it from published comprehensive-law counts. The error was corrected before publication, so it is a verification finding rather than a subscriber correction.
Scope of this record
These results describe two specific checks and their samples. They do not establish that every future error will be caught, and they are not a claim of legal review.