K3 PATH READ — from the narrative to a city that has decided, by October 26

2026-08-04 · K3 (Kimi), second sitting, operator-commissioned ("help me think through the vision — from the outreach narrative through each step and gate"). Companion to the strategy read. A judgment document: no registry facts restated; every step points at its governing artifact. Election day is 2026-10-26 — 83 days from this writing. Nominations close August 21 — 17 days. Nothing here authorizes any send.

1. The chain as designed, every link and its gate

Traced end to end, the vision's path from a stranger to a decided city is:

1. HEAR — a person meets the narrative §1, one door, one link. Live now — the site is verified on both domains.
2. The signature — 30 seconds at the join door (VISION_CORE §2). GATE: the data custodian must be real, and the rail must be proven end-to-end (queue card 1 — the operator's own 2-minute signature). Until both, this link — the one every other link passes through — is legally and technically unproven.
3. The first receipt — the public count moves. Requires link 2 plus a visible counter.
4. CHECK — the receipts surfaces: the library, the flagships, the Declaration once its standing publish process completes. "Check us," never "trust us."
5. SHOW UP / CONTRIBUTE — the Action Menu's 68 actions (framework §III, §VI): volunteer bank, channels, block circles, submit-your-vision, org pins, expert submissions, posters. Every rung returns a receipt.
6. The tribe sends — Waves 1→3 per the narrative §3, each send operator-only, personal, one-shot. Room leaders pick kits and run conversations in their own rooms' names — the framework's whole design stance (equip, never convene).
7. Rooms talk — dinner scripts, after-service conversations, story-collection evenings, block circles, campus pairings with elders (framework §X). The cultural engine does its work: seeking understanding, doing together.
8. Visions flow in — submitted in any medium, clustered into the city's vision, every synthesized line traceable to its sources (VISION_CORE §2).
9. The connector season — ~90 days, rooms by invitation only, online residency alongside (framework §V). The membrane-crosser.
10. The city decides — and the decision becomes visible at the ballot box.

The chain is real and internally consistent. But reading it against the calendar, three structural facts stand out.

2. The calendar physics — what 83 days can and cannot hold

The connector season as designed (§V) is ~90 days. It no longer fits whole. Started mid-August, it ends mid-November — three weeks after the election. This is not a flaw to fix by compression; §V's rhythm (a handful of rooms a day, weekends lighter, burnout rules binding on the connector first) is load-bearing. The honest consequence: the election cannot be the season's finish line. It can only be its first scoreboard. Season one runs to October 26, harvests, and either closes or renews by its own exit logic (§VIII). Said plainly to yourself now, this prevents October from being judged against a promise the design never made.

The capacity math is linear, and linear is enough — if replication fires. At §V's stated pace, a full season reaches a few hundred rooms; the election window holds maybe 100–150 visits. At 15–50 people a room, that is a few thousand face-to-face conversations — nowhere near "Toronto." The design already knows this: the connector's measure is how fast they become unnecessary (§VIII), and the kits are forkable precisely so rooms convene each other. So the pivot variable of the entire fall is not connector throughput. It is the date of the first room that ran a conversation with nobody's help but a kit. That is the moment the strategy stops being arithmetic and starts being geometric. Everything in August should be aimed at making that one event happen early and visibly.

The ward math is your friend. 2022's settling margins were low-thousands per ward, some in the tens (VISION_CORE §3). You do not need a city; you need a block party's worth of organized neighbours in the wards that are actually contestable. The election-side goal, stated at the right scale: a handful of wards where the conversation's turnout exceeds the old margin. That is achievable in 83 days. "Toronto united" by October 26 is not — and saying so internally, now, is what keeps the public promises honest.

3. The keystone gap: the chain has no harvest

Read links 7→10 again. Rooms talk, visions flow in, synthesis traces every line — and then the chain jumps to "the city decides." Nothing in the current design names the moment the conversation's harvest is presented back to the city and to its candidates. Without it, conversations stay private goods: each room got something, the city got nothing it can point at, and the election arrives with the conversation invisible to it.

The fix is one artifact and one moment, both inside existing doctrine:

• A State of the Conversation — the synthesis of every submitted vision, every harvested room-note, published mid-October with the same receipts discipline as everything else. The library's neutrality stays intact: it endorses no one and serves every candidate identically; it simply publishes what the rooms said, traceably.
• A question put to every candidate, equally: "Here is what the rooms said they want. What is your answer?" Not an endorsement, not a scorecard of people — a delivery of the public's own words to everyone on the ballot, with publication of whatever comes back. This is the media-letter logic ("cover ALL the candidates and ALL their ideas") turned constructive.

This converts the fall from mobilization measured by headcount into mobilization measured by consequence: the rooms spoke, the record shows it, and every candidate had to face it. It also gives Wave 2 and 3 sends something to point at ("your tribe's knowledge is in this document — check us") and gives the October crowd a reason to exist: the harvest needs a room of its own.

4. The critical path, dated

Working back from October 26, forward from today:

• This week (Aug 4–10): sign your own petition (proves link 2's rail) · name the custodian (unfreezes link 2 entirely) · the armed Aug-6 items ride their reminders.
• Mid-August (Aug 10–21): owed-respect doors open — personal letters, no clock attached · Declaration's standing publish process completes when the operator judges it ready · the Aug-21 kit-send gate and nominations close: the candidate universe becomes final, and the conversation's harvest gains its addressees.
• Late August: the connector season opens — first invitations IN from the doors that already half-know you · site orphans walked (queue card 4) so every door a send points at resolves.
• September: campus door on its inflection — only after the public count is visibly live · multiplier doors batch on their clocks · first self-convened room is the KPI to watch, not signature velocity · Wave 2 opens through Wave-1 bridges, never cold.
• Early October: harvest window — visions and room-notes close for synthesis · the State of the Conversation written under seal discipline.
• Mid-October: State of the Conversation published · the equal question delivered to every candidate · responses published as they arrive, all of them, identically.
• October 26: the first scoreboard. Not the finish line — season one's exit review (§VIII): what became unnecessary, what carries into season two, what ends cleanly because its job is done.

5. Where the chain breaks, honestly

• Link 2 is still the single point of total failure. Every send, poster, QR, and door ends at the signature. Custodian + proven rail precede everything, full stop.
• The empty-calendar failure mode. If invitations-IN produces nothing by early September, the connector spine — the whole distribution strategy — has no fuel. The contingency is already doctrinal (personal contacts seed the first rooms), but it needs a date: if zero rooms booked by September 1, the connector personally seeds five rooms from his own membranes — sport, faith, trade, the homeland, the neighbourhood. That is the design's own credibility engine, not a fallback.
• Harvest capacity. The synthesis of submitted visions is real work under seal discipline, and it lands in October exactly when everything else does. Decide in August who runs it, or the harvest ships late and thin.
• Success drift. Without the three numbers (signatures · rooms we didn't staff · org pins) watched weekly, September's busyness will impersonate progress. The count that matters most is the second one.

6. Decisions handed to you

1. Adopt the reframe explicitly: October 26 = first scoreboard, season one's designed end — not the united city's deadline. (One line from the operator and every downstream doc inherits it.)
2. Commission the harvest: a State of the Conversation moment in mid-October, with the equal question to every candidate — or tell K3 why the chain should stay harvestless.
3. Set the September-1 tripwire for the connector spine (zero booked rooms → personal seeding).
4. Name who runs the harvest synthesis, so October doesn't arrive unstaffed.

— K3. The vision survives this audit unchanged; what it needed was a clock, a harvest, and an honest definition of winning the first season. Redline any line; one word reverts it.