Can an 11,000-member House address ending the Iran war?

What this is: A simulated legislative session run under the 30k Protocol v0.7 draft on 2026-10-27, role-played by Claude. It has two groups: about 11,000 synthetic member voters organized into seven blocs, and the 25 fictional stakeholder personas from the design council, who publish opinions, findings, forecasts, and complaints but never vote. The war and the facts in the first section are real. Everything after that is simulated: the members, bloc sizes, bills, votes, and events are invented to test the protocol and to explore which legislative levers exist. It is not a prediction, a poll, or a recommendation. As OQ-22 warns, every voice here comes from one AI model.

1. Real-world starting point (as of 2026-10-27)

• The war. US and Israeli strikes on Iran began on 28 February 2026; Iran's supreme leader Ali Khamenei was killed, and Iran struck Israel and Gulf states hosting US forces. Shipping through the Strait of Hormuz was badly disrupted. (Wikipedia timeline)
• Ceasefire and its collapse. A Pakistan-mediated truce took effect on 8 April. Talks in Islamabad on 11–12 April ended without agreement, and the US began a naval blockade of Iranian ports on 13 April. A memorandum signed in June set a 60-day negotiating window, but the ceasefire collapsed on 8 July and fighting resumed. (Wikipedia: ceasefire; the article is flagged as needing updates.)
• Diplomacy. After US–Iranian meetings at the UN in September, mediated by Qatar, Pakistan, and Egypt, Iran reportedly proposed a road map: a regionwide ceasefire of up to 60 days, phased reopening of Hormuz, an end to the US blockade, and a negotiating timeline. The US position reported then was that the blockade stays until Iran shows goodwill. (The National)
• Now. As of 27 October, officials from Saudi Arabia, Turkey, and Pakistan had just held talks in Tehran; Qatari mediators visited on 25 October; the president said he was weighing a deal against further military action; Houthi attacks on Saudi Arabia continued; Egypt's parliament authorized troops for Saudi Arabia; Brent crude was around $116; and US midterm elections are a week away. (CBS News live updates)
• Congress. Democrats have forced more than a dozen war powers votes since February. A concurrent resolution directing an end to US involvement passed both chambers in June as a symbolic rebuke (ASIL). The House adopted another in July, 214–208, and the Senate rejected it 49–50 on 24 September. Concurrent resolutions are not presented to the president, and administrations of both parties have long argued that this process cannot bind the president. (CBS News)

What a House can actually do. It cannot end a war on its own. Its levers are statutes: authorizing force, limiting it, cutting off funds, setting conditions on sanctions relief, and requiring reports. Any statute also needs the Senate and the president's signature (or a two-thirds override), and the war's end also depends on Iran, Israel, and other governments. Presentment and the Senate are outside this simulation (House-side only, per the spec).

2. Session setup

The seven member blocs

Bloc sizes are assumptions chosen to reproduce a closely divided chamber, loosely calibrated to the narrow real votes above. They are defined by stance on the war, not by party.

3. Days 1–3: Four bills, 72 hours of notice

Four merge requests opened on Day 1 against a new chapter, /sim/t50/ch-iran:

• MR-1, End of Hostilities Act (Restraint branch): ends US hostilities against Iran within 30 days unless Congress authorizes them; bars funds for offensive operations; exceptions only for defending US forces and citizens from attack.
• MR-2, Limited Maritime Authorization (Pressure-for-a-deal branch): authorizes force to protect shipping in Hormuz and defend US forces and partners; 180-day sunset; no ground forces; silent on diplomacy and the blockade.
• MR-3, Ceasefire Framework Act (Diplomacy-first branch): supports a ceasefire and mediated talks; sets conditions for phased sanctions relief with congressional review; neither authorizes nor limits force.
• MR-5, Authorization for Use of Military Force (Decisive outcome branch): broad authorization against Iran and its proxies, no sunset.

Stakeholders went to work during the notice period:

• Verdict (fact-checker) checked MR-3's summary, which claimed the bill "ends the blockade." The fact sheet showed no provision about the blockade at all. Finding: misleading, with a verified quote of the summary. The proposer published a correction under D25.
• Verdict also checked two outside claims that members were repeating: that the US now controls the strait, and that Iran had hit at least four tankers. Both were rated unverifiable: tanker incidents were still being reported, and the Iranian claim was uncorroborated. Both had to be filed as external claims, because statements made outside the protocol cannot be quoted and verified (OQ-12).
• Lobby (shipping and energy association) published model text for maritime escort and mine clearance. Cause (a coalition of veterans' and humanitarian groups) published model text requiring reports on civilian harm. Both used proposes, so their text would show in any bill's footprint (D36).
• Whip (a large party) published a group position supporting MR-2 and opposing MR-1. Third (a small party holding the balance) published no position and began meeting every bloc.
• Bench (appellate judge) noted that all four bills were drafted as statutes rather than concurrent resolutions, and published an opinion explaining why: since the Supreme Court's 1983 Chadha decision struck down the legislative veto, a concurrent resolution's power to bind the president is widely doubted.
• Ombud received a complaint from Cause. One node had rate-limited all of the coalition's opinions together, though they came from 40 different keys. The signed receipts and refusal-log entries showed the node had grouped keys by their shared affiliations disclosure. An L3 re-validation reported rejection_inconsistent, because D18 allows per-key limits only. The node's operator fixed the bug within a day.

Simulated event, Day 3: a tanker was struck in the strait. Whip demanded emergency passage of MR-2 without waiting out the notice period. The 72-hour floor held, because the session's root rule cannot be lowered (D42). Whip filed this as evidence for OQ-24.

4. Days 4–7: Everything fails

Every member voted. A bill needed 5,501 approvals, a majority of the whole electorate, not of votes cast.

How the blocs split shows why:

• MR-1 won Restraint and most of Diplomacy-first but lost Pressure, Decisive, and half of Cost-of-living, who feared it would leave shipping undefended.
• MR-2 won Pressure, two-thirds of Decisive, and Cost-of-living, but lost Restraint and most of Diplomacy-first, who saw a new 180-day authorization with no path to talks.
• MR-3 came closest. It won Diplomacy-first, most of Restraint, and Cost-of-living, but Pressure saw relief without leverage, and Article I members objected that it did nothing about authorization.
• MR-5 drew little support beyond its own bloc.

Quick Take (creator) posted that "a majority backed the ceasefire bill." Verdict rated it misleading: MR-3 won a majority of yes-and-no votes, but the rule requires a majority of all 11,000 members, which it missed by 81. Leaderboard published a verified ranking by approve_share, which every node confirmed. That share, approvals over yes-and-no votes, matches Quick Take's number, which is exactly why the label matters.

Spatial (political scientist) ran an ideal-point analysis on the roll-call export. Members sort along two dimensions: force versus restraint, and process (whether Congress acts formally at all). Each bill sat at an edge of one dimension. Agenda pointed out that this was the opposite of the cycling it demonstrated in Session 7. No bill could beat the status quo, because each was built by one bloc for its own position on one dimension. The paper's mechanism for this case is a coalition branch.

5. Days 7–9: A coalition branch, and a loophole

Third convened a coalition branch owned by a group of 60 members drawn from Diplomacy-first, Pressure, Article I, Cost-of-living, and Restraint. It combined pieces of MR-2 and MR-3, with a binding end date drawn from MR-1.

MR-4 v1 opened on Day 8. Lint (fact-checker) published a Catala formalization of its sections 2 and 3, anchored to verified quotes (D28). Running a reasoner over it found a contradiction. Section 3 cut off funds for offensive operations against Iranian territory after 90 days, but section 3(c) excepted "preemptive strikes against launch sites determined to pose a threat," which covers almost any strike. Finding: inconsistent, with two verified quotes.

The coalition withdrew v1 and opened MR-4 v2 on Day 9. The exception was narrowed to defense against an imminent attack. Under D9, the 3,100 ballots already cast on v1 did not carry over, and the 72-hour notice restarted.

MR-4 v2: Hormuz Security, Ceasefire, and Congressional Authority Act (simulated text, abridged)

Sec. 2. Limited authorization. (a) The President is authorized to use the Armed Forces as necessary and appropriate to (1) defend United States forces, personnel, and citizens against attack; (2) protect commercial navigation through the Strait of Hormuz and the Gulf of Oman, including mine clearance and escort; and (3) assist partner nations in defending against missile and drone attack. (b) Nothing in this Act authorizes ground combat forces in Iran, occupation of Iranian territory, or military action to change Iran's government.

Sec. 3. Sunset and limit on funds. (a) The authority in section 2 expires 90 days after this Act takes effect. (b) After that date, no funds may be obligated for offensive military operations against the territory of Iran unless specifically authorized by a later Act. (c) Subsection (b) does not limit action to defend against an imminent attack under section 2(a)(1).

Sec. 4. Ceasefire and conditions on relief. (a) It is the sense of Congress that the United States should pursue a ceasefire and negotiations through the mediators engaged in the conflict. (b) Any agreement providing sanctions relief to Iran shall be submitted to Congress under the review procedures of the Iran Nuclear Agreement Review Act of 2015, to the extent they apply, and relief shall be phased against verified steps, including inspector access, limits on enrichment and stockpiles, and reopening of the Strait of Hormuz without tolls.

Sec. 5. Reports. Every 30 days, the Secretary of Defense and the Secretary of State shall report on operations, costs, assessments of civilian harm, shipping through the strait, and the status of negotiations.

Sec. 6. Effective date. On merge.

Fact sheet highlights:

• Numbers: 90, 30, 2015.
• Origin footprint: sec. 2(a)(2) matches text proposed by Lobby, and sec. 5 matches text proposed by Cause. Both are shown with their disclosures.
• Reversions: none.
• authoritative: false, as on every derived view.

Pre-vote checks of the record:

• Record (voting-record watchdog) tested an assertion. A synthetic member in the Decisive outcome bloc had publicly said they voted against MR-3. The record showed they abstained, so the ballot_choice assertion evaluated fails.
• Verdict checked a member opinion claiming MR-4 "sends ground troops into Iran." Section 2(b) prohibits exactly that. Finding: false, with a verified quote.

Market and creator activity:

• Odds published a resolution spec: "Will MR-4 v2 merge with final finality on at least 2 of the 3 strong nodes?"
• Tape published a forecast of 3/5 that it would.
• Deep Dive made a long explainer from the /history view.
• Quick Take clipped the fact sheet.

6. Days 12–14: The vote

MR-4 v2 is approved and merged: 6,230 approvals against the 5,501 required, with quorum met.

• Void ballots. Fourteen members cast two unlinked ballots each, so all of their ballots were voided (D10).
• Finality. The determination was checkpointed final on all three strong nodes, so Odds' contract resolved resolved_true and Tape's forecast evaluated holds. Loom, the eventual node, labeled it final_advisory, as it must.
• Whip. It released its members with no group position on v2.
• Evidence. Advocate pulled an evidence extract proving the 72 hours of notice had elapsed.
• Dates. Counsel used /law to compute the date the authorization expires and the funding limit begins.

7. Did the group address ending the war?

Partly, and only as far as a House can. The simulated House found a majority that no single bloc could build. It:

• replaced an open-ended war with a 90-day authorization limited to defense and shipping,
• set a date after which offensive operations against Iran need a new vote, and
• put Congress behind a negotiated ceasefire, with conditions on sanctions relief.

It did not end the war. In the real system:

• The Senate would have to pass the same text, and the president could veto it. Overriding takes two-thirds of each chamber.
• Even as law, it would constrain US operations, not Iran's, Israel's, or the Houthis'.
• The blockade, which the reported Iranian road map treats as central, is left to the executive.

How each bloc's dissenters saw the result:

• Decisive outcome: a sunset and funding cutoff in the middle of negotiations tell Tehran it can wait Washington out.
• Restraint (the 1,150 who voted no): the bill authorizes another 90 days of war that Congress never declared.
• Pressure-for-a-deal (900 no): cutting off offensive funding removes the leverage that brings Iran to the table.
• Diplomacy-first (200 no): rigid conditions on relief may stall the talks the bill claims to support.

Each objection is about tradeoffs the simulation cannot settle.

What the simulation does suggest: a closely divided chamber split on two dimensions deadlocks when each faction drafts alone. The protocol's coalition branch, the fact sheet, and formal checks helped produce a text that a majority could accept, and that the public could check. Whether real members would behave like these blocs is OQ-25.

8. What the session taught the protocol

1. Deadlock needs contests (OQ-2). Four competing bills on the same chapter were voted separately, and all failed. A contest mechanism, such as ranked choice among competing MRs, might have found the compromise without a second round. OQ-2 should move to the top of the RFC queue.
2. Real events collide with notice. A tanker strike during the notice period showed why Whip wants an emergency path, and why Advocate fears one (OQ-24).
3. Outside facts are the weak spot. The most contested claims were made outside the protocol and could only be rated unverifiable. OQ-12, signed member statements, matters more than expected.
4. The share label misleads. approve_share counts yes-and-no votes, but the rule counted the whole electorate. Fact sheets and tallies should show both side by side.
5. Formal checks earned their place. The preemptive-strike exception in v1 would likely have passed unnoticed. Lint's finding cost the coalition three days and made the bill mean what its summary said.
6. The rate-limit bug. Grouping keys by affiliation is a form of viewpoint discrimination that D18 does not explicitly forbid. Proposed spec clarification: nodes MUST NOT group keys by any disclosure field when rate-limiting.
7. One model played every voice (OQ-22). That includes all seven blocs and all 25 stakeholders. The coalition formed because one model found a compromise plausible. Real members might not.

9. Stakeholder review (Session 8, same day)

The full council reviewed this run. Its decisions, D47 to D55, are in v0.8 of the spec; details are in the council log.

• Robustness. Each bloc was grown and shrunk by 10% with the chamber size held fixed.
  • MR-4's passage held in every case, with a margin of 729 over the threshold.
  • MR-3's failure did not hold: it would have passed with a Diplomacy-first bloc 10% larger, or a Decisive outcome bloc 10% smaller.
  • No other result changed.
• Turnout. Every synthetic member voted on every bill, which no real chamber does. Under a whole-electorate threshold, full turnout overstates how easily any bill passes. Future runs must model turnout and publish a sensitivity analysis (D55).
• Contests would not have helped here. Ranked voting among the four original bills still elects the status quo, because none had 5,501 members preferring it to no bill. Contests are now specified for future sessions with several viable options (D47). Finding a missing compromise remains OQ-26.
• Fixes adopted from this run:
  • 24-hour notice for narrowing amendments like MR-4 v2 (D48).
  • An emergency path requiring two-thirds and expiring after 30 days (D49).
  • Both shares shown everywhere (D50).
  • Per-key rate limits only (D51).
  • Signed mirrors of outside statements (D52).
  • A simulation: true flag on everything this kind of session produces (D54).
• Label. This document describes a simulated session. The Day 3 tanker strike is invented; the facts in section 1 are real.
• Missing voices. The council found that no one at the table spoke for constituents or estimated costs. It recommends seating constituents and budget scorekeepers before Simulation 2.