Pending orders and protective execution records
Keep submissions, acknowledgements, cancellations and fills distinct when investigating an evaluation event.
An order is a lifecycle
A pending order is not a completed trade, and a click is not a fill. Start the record with the order identity and collect the submission, acceptance, modification, cancellation request, acknowledgement and any execution as separate observations. Not every source exposes every stage. State which stages are absent. This is especially important around a restricted event or session boundary, where the order state can change after the user action. The publication does not reconstruct hidden broker activity from a screenshot. It explains a record design that can preserve the known sequence without inventing the intermediate events needed for a favorable story.
Protective orders still produce executions
Stops and targets can trigger independently of a new manual entry. A preparation plan that checks only entry times leaves those events unexamined. Record the actual action type in the observation log, rather than labelling every event a trade. The news practice packet includes a stop-fill event to illustrate a scope match, not to recommend removing protective orders. Do not weaken native risk controls to avoid a calendar classification. An administrative question about execution permissions is separate from a strategy's protective policy. If those requirements conflict, preserve the conflict for authoritative review rather than pretending the worksheet has solved it.
Cancellation requires evidence
A cancellation request and a confirmed cancellation are different. The declared-window checker's cancel action is only an opaque supplied event; it does not certify that an order was no longer active. If your source provides both request and acknowledgement timestamps, keep both in the detailed journal. If it provides only one, specify which one. An invented example where a request precedes a deadline cannot prove that an actual order was absent afterward. Avoid closing a dispute with a screenshot that shows only the request. The relevant account owner and its authoritative order record determine what the platform recognized.
Keep partial fills and replacements separate
One order identity may produce several executions, while a replacement may have a new identity. Do not collapse these into a single convenient timestamp. A partial close can leave residual exposure, and an amended quantity can change the sequence without changing the headline order name. Store each observed fill under its own event identifier and retain the link to the parent order. This version of the public checker does not calculate quantities or net exposure; it compares only declared execution events and windows. Its event count therefore is not a trade count, a position count or proof that an account became flat.
Respect the available resolution
Whole-second timestamps cannot resolve the order of events inside the same second. The checker preserves input order for tied times without claiming that order was observed. A millisecond source should remain intact in a richer private investigation rather than being silently truncated for convenience. The public sample uses whole seconds as an explicit teaching convention. A result at an inclusive endpoint is a result under that convention. It does not establish a programme's timestamp tolerance. Preserve time precision, source timezone and clock ownership alongside the raw record when an actual boundary question depends on them.
Export observations, not a verdict
A complete investigation packet contains the account card, relevant calendar source, raw event identities, resolution, missing stages and the resulting interpretation. A match produced by this site is a reproducible classification of supplied inputs. It is not a firm's violation finding, a payout decision or an independent audit of executions. Keep approvals null until supported by the authoritative process outside the tool. A canceled or unmatched event can still be subject to another rule. This boundary makes the journal useful: it identifies exactly which question the available evidence can answer, and which question requires a different record or responsible decision maker.