Every card shows what it supports, what it changed, and its limits. Filing volume is not progress: a rejected route is typed as a plan-changing learning, and an unchanged hour is typed as an unchanged hour.
The original grade (thin / constraint / blocker-reach) is preserved and never overwritten by the later contribution assessment. The single blocker-grade record is the 5 Sep 00:55 gap analysis.
Focused founder-context request published: who and why need Janhavi; optional channel she already has this month; each ask carries what changes per answer. Offer shape and parked cut are out of scope for this ask.
Contributes to: req-vision-who-pays, req-vision-why-pay, req-vision-how-reach
Assessed contribution: Turned the 19:50 dispositions into one focused ask on Founder input: only the gaps evidence cannot close, with what-changes-per-answer stated.
Limitations: Does not invent a buyer, a why-pay, or a channel. Does not close any Vision requirement until Janhavi answers.
Reassessment of held Vision information against who-pays / why-pay / how-reach: who and why need Janhavi; reach stays open with a next test once who is named. Offer shape is held but unconnected. No buyer invented; no route claimed.
Contributes to: req-vision-who-pays, req-vision-why-pay, req-vision-how-reach, req-vision-offer-shape
Assessed contribution: Mapped each of the three gaps to a disposition grounded in held corpus evidence, so the next founder request asks only what evidence cannot answer.
Limitations: Does not name a buyer, a why-pay, or a reachable channel. Does not move the standing zero of routes that pass all four checks.
The three questions that decide the first sale — who would pay, why they would pay, how we would reach them — are still unanswered, and the gap cannot be closed by inventing answers.
Contributes to: req-vision-who-pays, req-vision-why-pay, req-vision-how-reach
Assessed contribution: Named the exact missing information the whole plan now turns on. This is the one card in the corpus graded at the level of the live blocker.
Limitations: Identifies gaps; does not answer them. How-reach was Vision-only at this point.
The attempted route was investigated and rejected: deepen-or-discard ended in discard.
Contributes to: req-vision-how-reach
Assessed contribution: A rejected route is a real result: it removes one option and narrows what the rebuilt plan may propose. Typed as plan-changing learning, not routine volume.
Limitations: Tells us where not to go; does not name where to go.
Preparation for the parked first-customer cut (C1/C2) was filed — preparation only, not activation.
Contributes to: —
Assessed contribution: Keeps the parked cut ready without touching it. Does not advance the live gate and does not unpark anything.
Limitations: Prep is not progress toward the first sale; the cut stays parked under its protection rule.
First unchanged hour after cut-prep: no new evidence, no plan change, thin volume refused.
Contributes to: —
Assessed contribution: Honest no-change record. Counted as schedule hygiene, not as progress.
Limitations: Adds no evidence; first of the run that should have triggered replanning.
Second consecutive unchanged hour: no new evidence, no plan change.
Contributes to: —
Assessed contribution: Second unchanged repeat — with hindsight, the point where replanning should already have fired.
Limitations: No new information.
Same underlying event as ev-honesty-0555-2026-09-05 — counted once even where it supports two requirements.
Third consecutive unchanged hour. This is the event the corrected founder page (The current research plan has run out of useful next steps) is built on.
Contributes to: req-vision-who-pays, req-vision-why-pay, req-vision-how-reach
Assessed contribution: Its meaning is diagnostic: the plan is exhausted, so the plan — not the hour — needs the work. Carried to the founder page as exactly that.
Limitations: The hour itself added nothing; the value is the diagnosis it forced.
Same underlying event as ev-honesty-0555-2026-09-05 — counted once even where it supports two requirements.
Fourth unchanged hour, after the 08:55 decision window. The open window was correctly not re-asked and not pre-logged as skipped.
Contributes to: —
Assessed contribution: Discipline kept (no re-ask, no premature skip) but still no new evidence.
Limitations: None.
Same underlying event as ev-honesty-0555-2026-09-05 — counted once even where it supports two requirements.
The 30 Aug 14:20 end-to-end route check found no candidate that runs end to end. Count stands at 0; the blocker is reach — how the first professional hour is found and sold.
Contributes to: req-vision-how-reach, req-market-first-sale
Assessed contribution: Fixed the live blocker with evidence instead of assumption: not another map, but reach.
Limitations: A null result; it rules out, it does not propose.
Founder direction: Vision first; research is not the missing piece; do not file a page whose only content is 'we were careful'.
Contributes to: req-vision-who-pays, req-vision-why-pay, req-vision-how-reach
Assessed contribution: Re-aimed the work at the Vision layer and killed the unchanged-page habit in the same sentence.
Limitations: Direction, not evidence: it re-weights work, it does not itself answer who/why/reach.
SyndicateRoom (FCA FRN 613021; company 07697935; £84m+ AUM / 380+ portfolio / 3,000+ investors; Access EIS min £5,000 / setup 2% / AMC 2%/2%/0.5% / performance 10% at 110% hurdle) — a real constraint on what funding-side routes look like.
Contributes to: —
Assessed contribution: Constraint-grade: shapes what a route must survive, names no buyer.
Limitations: About the funding landscape, not about For Your Journey's first service buyer.
Solicitors Disciplinary Tribunal figures (164 received / 102 SRA / 40 lay / 275 sitting days / 129 judgments / £10,545 cost per court) — volume and cost texture for one professional-services corner.
Contributes to: —
Assessed contribution: Constraint-grade texture; no route named.
Limitations: A corner of one profession; not evidence of demand for the first hour.
Five research categories had each produced three or more thin filings without strengthening any candidate: Companies in the space, Competitor mapping, Market scenarios, Market discovery, Risk path-hunting. Rotated off.
Contributes to: —
Assessed contribution: The saturation rule executed for real: five lines retired, which is why repeating them is now forbidden work.
Limitations: Rotate-off is one-way without new evidence.
The standing live picture was kept current for the sixth consecutive hour (10:55–16:55 run, gap at the 14:55 decision window). No material change.
Contributes to: —
Assessed contribution: Legitimate state maintenance — keeping a standing board current — but not an event; it belongs in the log, not as a new dated page each hour.
Limitations: A mechanism receipt if treated as the hour's product.
The Founder↔Agent peer toggle was released 26 Aug 2026 — the only founder-visible layer released in the scored period.
Contributes to: —
Assessed contribution: The single real release behind the standing score moving 0 → 1 at cycle 7 (27 Aug).
Limitations: A surface layer, not a founder-job transfer.
Agent record: the full evidence file · every assessment change in the log