Sources
The Sources tab — Sources — the audit trail. — is where the scan shows its working. Every item in the loaded snapshot is traced to the document behind it, and the page lays out what was searched, how citations were matched, what was verified, and how the scan was configured.
It exists for the moment trust is questioned: before you forward a briefing to a regulator-facing committee, when a number looks surprising, or when someone asks “how do we know this?”. The hero’s meta block sets the tone immediately — the engine the scan ran on, the citation-audit verdict, and the tally of checked claims across every surface — the same totals as the scorecard below: CITATION AUDIT ● COMPLETE over 112 VERIFIED · 19 LOCATED · 7 UNSOURCED.
On this tab the ticker strip swaps to trust statistics: ITEMS · JURISDICTIONS · SOURCED · PRIMARY — how many checked claims have a source, and the share of verified claims resting on a primary regulator’s own page. Segments the snapshot cannot support are omitted rather than shown as zero — an absent number is honest, a fabricated 0 is not, and the page holds that line throughout.
The Trust Scorecard — The trust numbers.
The scorecard is the one-panel summary, with the citation audit’s status in its header (green Citation audit complete; amber for partial, failed or timed out; muted when the audit was not run — with a count such as 1 of 84 unchecked when some cited sentences could not be checked). Two cards sit side by side:
| Card | What it measures |
|---|---|
| Claims checked · every surface | Of the factual claims in the narrative, the enforcement priorities and the item summaries, how many are sourced — verified plus located — for example 131 of 138 claims sourced, split into verified, located and no source found. Analysis sentences are counted separately underneath; they are our own reading and are not checked by design. |
| Verified claims by page tier | Of the verified claims, the share resting on a primary regulator’s own page — for example 85% on primary regulator pages — with the split across primary regulator, supranational and secondary pages. Each claim counts once, by the strongest page that verified it. |
The two cards answer different questions: the first says how much of the briefing a page actually supports; the second says who those pages are. A scan can be fully sourced on secondary commentary, and the second card is where that shows.
Source provenance — Evidence for each claim.
The main list follows the briefing’s own order:
- The briefing — the executive narrative, with every page its claims were checked on.
- Enforcement priorities — one row per priority.
- Items, grouped by jurisdiction, each group noting how many of its items rest on primary pages.
Each row carries:
- The title — for an item, click it to open the shared detail panel.
- An evidence chip — the strongest tier among the pages that verified one of its claims: Verified on primary regulator pages, Verified on supranational pages, Verified on secondary sources only, or Not verified. A page that only located a claim does not count.
- A tally — for example
6 claims · 5 verified · 1 located. - Every page the claims were checked on — the host, its tier (Primary regulator, Supranational, Secondary), the page title as a link, and how many claims it verified or located. The primary document — the primary regulator’s page that verified the most claims — is marked
↗ PRIMARY DOCUMENT. - For items, status, impact and topic pills.
An item the audit could not check says so — Not checked — the audit could not check this item; the source shown is the Analyst’s own — and keeps the Analyst’s source link.
Snapshots from before the evidence-based audit keep their earlier layout: a sourcing-confidence chip on each item, a ✓/≈/? cross-reference badge on its source URL, a scorecard of editorial claim coverage, confidence mix, cross-reference mix and tunnel substitutions, and a Source cross-reference worksheet under Analysis. Those signals are no longer produced for new scans.
Analysis — the working
Beneath provenance sit the scan’s collapsible worksheets:
- Search queries & sources — the actual queries issued, turn by turn, with each turn’s retrieved source URLs expandable underneath.
- Citation matching —
x/y ITEMS ATTRIBUTED, with per-item scored source matches. When items could not be attributed above threshold, an amber band says so rather than hiding it. - Relevance filtering — appears only on older snapshots that used a retrieval-filtering step; current scans do not produce it.
Scan configuration — How this scan was run.
The configuration panel records the run itself: jurisdictions, topic areas, the engine, the time horizon, entity types and any custom topics, followed by one row per analysis stage that ran — including whether the verification and citation-audit stages were enabled for this scan, and which translation languages were requested. The foot line stamps SCANNED date and time, DURATION, SEARCHES, and the scan’s cost (plus a running total of any export costs for this snapshot).
On sample data this panel says “Run a live scan to see configuration details here.”
Methodology — the key
The final panel, Methodology — how to read the trust signals, is the in-product key to everything above: the evidence chips and what counts towards them, the Primary document, each result a citation number can show (Verified, Source located. Check wording, No source found, Analysis — and Not checked, when the snapshot has one), and the amber audit chip.
Its most important point is worth restating here: the audit checks the briefing against the pages this scan read — it does not check those pages against the world. A Verified claim is one a page states. Citation audit partial means some cited sentences could not be checked — a whole section, or more than one in ten — and they are marked in the text; Citation audit failed means none could be. Neither means the scan failed, nor that any finding is wrong.
Do not read absence as failure on historical snapshots. Scorecard cards, worksheets and chips each appear only when that snapshot recorded the underlying data — a snapshot from before a given audit capability existed will show fewer, or different, signals, honestly. Compare trust signals only between snapshots of a similar vintage.
Worked example: standing up a number
The Briefing claims a licensing figure you are about to quote to your board.
- On the Briefing, open the sentence’s citation number. The claim with the figure reads Verified, quoting the page’s own words, on an
esma.europa.eupage tagged Primary regulator. - Open More → Sources. The ticker reads
SOURCED 131/138 · PRIMARY 85%— the audit ran, and nearly every checked claim has a source. - Under Evidence for each claim, find the item. Chip: Verified on primary regulator pages; the ESMA page is marked
↗ PRIMARY DOCUMENTand verified four of the item’s claims. Open it and read the figure in the regulator’s own text. - If a neighbouring claim reads Source located. Check wording, read that page too before relying on its detail.
Now you can quote the figure with its provenance attached — the regulator’s own page, read by this scan, stating the figure.
A non-obvious tip
Read the two scorecard cards together. A high claims sourced count with a low primary share means the briefing is well supported, but mostly by commentary — common on fast-moving topics where the regulator has not yet published, and a cue to budget time for checking the primary texts as they appear. The per-item chips under Evidence for each claim show exactly which items those are.
Where next
- Overview — the evidence chip, tally and primary document resurface on every detail card.
- Briefing — where the citation numbers live.
- Scans and snapshots — how a scan is scoped and recorded.