How the monitoring works. Nothing hidden. Everything auditable.
EdgeLetter documents the assumptions underneath a decision, attaches public evidence to each one, and monitors that evidence over a defined review window. When the evidence changes, the change is recorded with its source and its timestamp. Every step of that process is documented here. If the archive says something was corroborated, this page explains exactly what that means.
Why publish this at all? Because a system that does not explain how it documents its own outcomes cannot be trusted. Every claim on this page corresponds directly to how the pipeline actually operates. If something in the archive does not match this methodology, that is a bug — not a feature.
01 / Detection and Structuring
How a development becomes a monitored assumption.
The pipeline runs continuously, reading public sources across geopolitical, trade, energy and macroeconomic domains. A development is only useful to this system once it has been broken down into the specific assumptions it bears on. That structuring step is what the rest of the process depends on.
1
Source ingestion
The pipeline reads public sources across geopolitical developments, military activity, trade and sanctions, energy and commodities, monetary policy, and corporate disclosure. Sources are read continuously rather than on a scheduled crawl.
2
Extraction and structuring
Each development is extracted into a fixed structure: what happened, what it bears on, which assumptions it touches, the mechanism connecting them, and the conditions under which those assumptions would no longer hold. That structure is preserved in the Protection Card.
3
Evidence strength assessment
Each candidate is assessed 0–100 for evidence strength. The measure reflects source quality, corroboration, and how specifically the development bears on a stated assumption. It is a measure of how well the claim is evidenced, not a view on any market or instrument. Only items scoring 55 or above are recorded.
RECORDING THRESHOLD: 55/100 Below 55 → held for further evidence, or discarded 55 and above → recorded as a Protection Card
4
Recording with timestamp
When a card is recorded, its timestamp is stored permanently. It cannot be changed afterwards. The timestamp is what makes the archive auditable — it fixes what was known, and when it was known.
5
Subscriber notification
Changes affecting a subscriber's recorded assumptions are delivered by email, in the dashboard, and by the other channels enabled on their plan. Delivery channels differ by tier. The speed and depth of monitoring do not.
02 / Evidence Strength
What the 0–100 measure means, and what it does not.
Evidence strength describes how well a claim is supported by the sources available at the time it was assessed. It is not a probability, not a rating of any company or instrument, and not a view on price. A high score means the evidence is well corroborated and specific. A low score means it is thin, vague, or single-sourced.
Score
State
What it means
72 – 100
RECORDED · WELL EVIDENCED
Multiple independent sources, specific and dateable claim. Recorded and monitored through its review window.
55 – 71
RECORDED · ADEQUATELY EVIDENCED
Meets the recording threshold. Evidence is sufficient to document, and the card notes where it is thinner.
40 – 54
HELD · INSUFFICIENT
Below the recording threshold. Kept under observation. Recorded only if the evidence strengthens.
0 – 39
DISCARDED
Too thin, too vague, or contradicted. Not recorded.
This measure gates what enters the archive. It is a documentation-quality control, applied before anything is recorded. It is not shown to subscribers as a live indicator, and it is deliberately not presented as a moving number attached to any position or instrument.
03 / Corroboration Doctrine — RECOG_v3
Exactly what "corroborated" means. No exceptions.
Corroboration is the most important word in the EdgeLetter archive. When a Protection Card is marked corroborated, it means one specific thing, defined by the RECOG_v3 doctrine. Not "widely reported." Not "broadly believed." The exact conditions below, and nothing else.
RECOG_v3 Corroboration Formula
CORROBORATED= second independent corroborating source ANDEvent AND ( ThemeORStrongLocation ) AND ( ActorORStrongLocation )
What each term means:
Event — A specific, dateable occurrence must be confirmed. "Tensions are rising" is not an Event. "Iran reimposed vessel restrictions on 18 April" is an Event.
Theme — A domain classification: Energy & Commodities, Geopolitical Risk, Trade & Sanctions, Defence & Security, and so on. Named strategic chokepoints satisfy Theme when paired with an Event.
StrongLocation — A named geographic chokepoint with established strategic significance. These satisfy either the Theme or Actor requirement when an Event is confirmed.
Actor — A named state, institution, or non-state actor whose actions are confirmed by the second source.
Second independent source — The corroborating source must be editorially independent of the first. Two articles from the same publisher do not satisfy this requirement.
Named chokepoints that qualify as StrongLocation:
Strait of HormuzSuez CanalRed SeaBab-el-MandebStrait of MalaccaTaiwan StraitSouth China SeaBlack SeaBosphorus
04 / Review Windows
How long an assumption is monitored before its outcome is documented.
Every recorded card opens a review window. During that window the pipeline continues reading public sources for anything bearing on the same assumptions. When the window closes, the outcome is documented against what the evidence actually showed — not against what anyone expected.
1
Window opens at recording
The window starts at the card's recorded UTC timestamp. Its length is set by the nature of the assumption — some conditions resolve within days, others need months. The due date is stored on the card and is visible before the outcome is known.
2
Continuous evidence collection
Throughout the window, further public sources bearing on the same assumptions are attached to the card. A second independent source satisfying RECOG_v3 marks the card corroborated. Contradicting evidence is attached in the same way and is never discarded.
3
Outcome documented at close
When the window closes, the system documents whether the assumptions held, whether conditions changed, or whether the evidence was inconclusive. This is automated. There is no editorial step and no opportunity to revise.
4
Evidence links preserved
Every card stores the URLs of the sources used. Anyone can open a card and check the sources for themselves. If a link later goes dead, the timestamp and the domain remain on record so the documentation can still be independently assessed.
05 / Outcome Documentation
How outcomes are recorded — including where nothing could be concluded.
After a review window closes, the system records what the evidence showed. This is automated and not subject to editorial judgement. Outcomes are published permanently and cannot be changed. EdgeLetter does not publish an aggregate success rate, because the purpose of the record is to document what was monitored, not to argue that the system was right.
HELD
Assumptions still supported
At the close of the review window, the evidence attached to the card continued to support the assumptions it documented. Nothing in the collected sources contradicted them.
CONDITIONS CHANGED
The underlying conditions moved
A later development changed the conditions the card rested on — a policy reversal, a ceasefire, a regulatory shift. The change is documented with its sources and the time it was observed, so anyone relying on the original assumptions has a record showing they no longer hold.
INCONCLUSIVE
Insufficient evidence to document an outcome
The window closed without enough independent evidence to say whether the assumptions held or changed. The card is preserved with this outcome rather than forced into a binary. Most cards in a monitoring archive end here, and that is normal.
NOT APPLICABLE
Outcome documentation not relevant
Applied to structural or continuing situations where a discrete outcome would misrepresent what was being monitored.
Why there is no accuracy percentage. EdgeLetter does not forecast, so there is nothing to be accurate about. The system documents assumptions and records what the evidence subsequently showed. A headline success rate would imply a predictive claim the product does not make, and would say nothing about whether the monitoring was useful to any particular decision.
06 / The Archive
What goes in, what can never be changed, and why.
The Evidence Archive is the permanent record of every Protection Card EdgeLetter has recorded. The rules governing what can and cannot happen to a card are not policies — they are how the system is built.
🔒
Protection Cards are immutable after freezing
Once a card is frozen, its headline, timestamp, evidence links, and documented outcome cannot be edited. The system stores them as insert-only records. There is no update path.
👁
All cards are publicly visible
Every Protection Card — including inconclusive ones — is publicly accessible at edgeletter.pro/proof-archive. No sign-in is needed to audit the complete record. Anyone can check.
⏳
Timestamps are recorded at the time, not retroactively
A card's timestamp is stored the moment it crosses the recording threshold. It is not assigned after corroboration, after the event resolves, or after any later review. The timestamp is the timestamp.
📊
Evidence links are stored with each card
The sources used are stored on every card. If a link goes dead, the timestamp and the domain remain on record, so the documentation can still be independently assessed.
❌
Nothing is removed
The archive contains the cards where nothing could be concluded alongside the ones where something clearly changed. This is not a marketing decision — it is a design requirement. A selective archive proves nothing. A complete one, including the unremarkable entries, is the only kind that can be trusted.
The methodology is only as credible as the archive that documents it.
Every claim on this page corresponds to a real mechanism in the pipeline. The best way to check it is to open the archive, pick any Protection Card, and read the evidence sources for yourself.