3,330 single-source entries
70.4% of legacy entries currently cite one source. A single direct primary record may be sufficient for a narrow formal fact; otherwise these records enter the remediation queue.
Generated QA inputs updated 2026-10-02T16:03:30Z
Archive quality dashboard
Editorial news coverage remains checked through 2026-10-02 5:57 PM EDT; archive-network measurements were checked through 2026-10-02T16:03:30Z, and legacy revisions are recorded through 2026-09-28. These totals are generated from canonical JSON. “Legacy-unreviewed” does not mean false; it means the record has not yet been revalidated under the stronger current-layer standard. Missing interpretive layers are measured as work to do, never silently invented.
5,064active entries rendered in the full archive
7,694attached source references
6,300distinct source URLs
4,716active legacy entries awaiting completed current-standard review
3,193active legacy entries awaiting Maybe / Therefore
11retired duplicate tombstones excluded from totals
Coverage continuity
The canonical legacy layer reaches 2026-08-01 and the generated national bridge begins 2026-07-13. The layers overlap, so the generated metric detects no inter-layer continuity gap.
Review receipt coverage
| Distinct measure | Count |
|---|---|
| Current entries with a content-bound claim-level receipt | 214 |
| Current entries retaining only their prior review-state label | 123 |
| Preserved material-development records (not additional events) | 425 |
Existing review labels were not converted into new inspection claims. New or changed content must carry source mappings and a review tied to its exact version. Inspect the ledger. Text and media inspection coverage for the retained feed is recorded separately in metrics JSON.
Evidence health
What the generated audit found
Source presence and source sufficiency are different controls.
1,360 weak-link-only entries
28.7% currently rely only on publisher homepages, search results, or query-result pages rather than direct supporting documents.
884 first-priority records
These entries combine a single citation with a low-specificity destination. They are the first automated remediation queue—not an allegation that every underlying event is wrong.
352 legacy source domains
5,329 unique URLs appear across 6,714 legacy source references. Repeated links remain visible rather than inflated into “unique sources.”
0 unresolved exact-text duplicate groups
Confirmed duplicates are preserved as redirecting tombstones and removed from active totals. Candidate detection is automated; merging remains an editorial action because records can share wording while documenting different lifecycle stages.
337 structured current records
92 are presently single-source; 0 lack a substantive Maybe / Therefore field; 337 carry an explicit reviewed or corrected state and 0 remain pending. New federated promotions cannot pass without both the reasoning layer and explicit review state.
1,538 Maybe / Therefore layers
32.5% of the legacy body already contains the competing-frame layer. The remaining 3,193 entries are a measured editorial backlog; federation publication now requires this layer.
3 same-date heading clusters
Stable IDs, exact fingerprints, normalized date/title headings, origin IDs, canonical targets, and lifecycle relationships are checked before import. Candidates are linked for review rather than automatically copied or merged.
33 repeated-heading review groups
These cross-date candidates are queued for adjudication. Repeated headings can represent recurring posts or distinct lifecycle stages, so they are not silently deleted or excluded until a revision names the surviving record or records the distinct-stage decision.
155 logged legacy revisions
Every applied correction or mechanical cleanup is recorded in the append-only revision ledger and guarded by its expected prior value.
1 normalized external crosslinks
These are counted separately from canonical entries. No outside item becomes a finding until its evidence, full reasoning structure, provenance, and deduplication checks pass.
Coverage by era
Active canonical legacy body
data/legacy_entries.json stores 4,742 rows: 4,731 active records plus 11 retained tombstones. The era counts below include active records only.
| Era | Entries | Review state |
|---|---|---|
| Campaign 1 | 526 | Legacy — not yet revalidated under the current standard |
| Campaign 2 | 627 | Legacy — not yet revalidated under the current standard |
| Formation | 352 | Legacy — not yet revalidated under the current standard |
| Post-presidency | 465 | Legacy — not yet revalidated under the current standard |
| Term 1 | 1,638 | Legacy — not yet revalidated under the current standard |
| Term 2 | 1,123 | Legacy — not yet revalidated under the current standard |
Definitions and limits
Source reference means one source object attached to one entry; repeated URLs count repeatedly. Distinct URL means a unique URL exactly as stored; redirects are not yet collapsed. Maybe / Therefore names the strongest plausible defense or uncertainty, then states the evidence-bound consequence, test, or remaining gap. Automated checks identify risk and inconsistency, while human review determines whether an entry is correct, sufficiently sourced, duplicated, corrected, or superseded.
Separate legacy derivatives
The Politicians detail index and companion DOCX/PDF files are frozen or separately generated legacy derivatives. They are not included in the canonical timeline totals and must not be presented as synchronized copies until rebuilt around stable record IDs.