Agent Swarm Domination | SREIC v2.0 and Real Estate Swarm v3.0

SECTION 00

What this is

Two designs, one demo. v3.0 is a five-colony hunter that scores distressed property, filters it, and contacts owners on four channels at once. v2.0, called SREIC or the Dominion, scales that idea to seven legions, adds capital formation, development, exits, counter-intelligence, tokenization, and a self-improving meta legion.

v3.0 hunt loop

Acquire targets, filter by distress score at least 7 and equity at least 30 percent, execute on at most 10 targets per cycle per colony, log the contact, evolve the colony. Hunt every 5 minutes: 288 cycles per day.

v2.0 Dominion loop

Seven legions act as one organism. Acquisition finds, Capital funds, Development forces appreciation, Exit extracts, Counter-Intel watches competitors, Tokenization fractionalizes, Meta-Architects breed and mutate the swarm every hour.

What the dashboard added

The zip dashboard adds live-style monitoring: legion cards, a US map with 100 map agents across 48 cities, a 24-hour performance chart, an activity feed, a command console, and spawn controls. All of it is simulated data in the source app.

Source claim, not a verified fact: the pitch deck states $3.8T in annual waste, a $42T TAM, 133x ROI, and a response time under 2 minutes against a 72-hour industry average. No independent verification is in either source.

Figure 1. The hunt loop as specified in v3.0, extended by the v2.0 Dominion evolution loop.

SECTION 01

Legion explorer

The seven legions from the Dominion interface. Counts sum to 5,150. Active counts are the dashboard initial values in App.tsx and sum to 4,822, which is 93.63 percent of the design target. Select a row to inspect it. Spawn controls are simulated and change the counts on this page only.

Legion	Design count	Active (source dashboard)	Share	Function

Acquisition Legion

Spawn adds to the selected legion only. Mass deploy adds 100 to each of the 7 legions, 700 in total, matching the dashboard log line.

Figure 2. Legion size, design counts from the Dominion interface. Bar length is proportional to count at 0.2 px per agent.

SECTION 02

US swarm map

Schematic map using the 48 city coordinates from the zip dashboard App.tsx. The source map is a simplified SVG, not a real geographic outline, and its 100 map agents are placed at random among these cities. This page plots one dot per city, deterministically. Select a city to see it.

Figure 3. The 48 cities in the source dashboard city list. Dots are clickable.

Selected city: None selected. Choose a dot.

Cities: Seattle, Portland, San Francisco, Los Angeles, San Diego, Phoenix, Las Vegas, Salt Lake City, Denver, Albuquerque, Dallas, Houston, Austin, Oklahoma City, Kansas City, Minneapolis, Chicago, Detroit, Indianapolis, Columbus, Atlanta, Miami, Tampa, Charlotte, Nashville, Memphis, New Orleans, Boston, New York, Philadelphia, Baltimore, Washington DC, Richmond, Raleigh, Pittsburgh, Cleveland, St. Louis, Milwaukee, Des Moines, Omaha, Boise, Billings, Cheyenne, Fargo, Sioux Falls, Little Rock, Jackson, Birmingham.

SECTION 03

Hunt verticals explorer

The five v3.0 colonies. Each has its own signal source in the code, its own distress logic, and the same filter: distress score at least 7 and equity at least 30 percent, then at most 10 targets per cycle per colony.

Vertical	Signal in source	Distress logic	Base score in code
Probate	PropStream probate, counties Maricopa AZ, Harris TX, Miami-Dade FL, Clark NV; last 7 days	Fixed at 9. Death is treated as maximum distress in the code.	9
Foreclosure	RealtyTrac pre-foreclosure, FL, TX, AZ, CA; equity minimum 20 in the query	min(10, 5 + (30 minus days to auction) / 3)	Formula, see calculator
Divorce	Trellis court records, Cook IL, LA CA, Harris TX; filed after 2025-01-01; real property only	Fixed at 8; equity assumed 50	8
Absentee	PropStream absentee, FL and TX; eviction within 90 days; equity minimum 30	Fixed at 7	7
Code violation	Municipal code enforcement, Phoenix, Houston, Miami; open, fines at least $1,000, last 30 days	min(10, 6 + total fines / 1,000)	Formula, see calculator

Select a vertical to explore

Source claim, not a verified fact: the code assumes equity values in several places (divorce 50 percent, code violation 40 percent) rather than reading them from a record. Those assumptions are reproduced here and labeled.

Simulated hunt cycle

Run one cycle with the current vertical and legion. Nothing leaves this page. The feed is deterministic text generated from your selections.

[SYSTEM] Swarm demo loaded. No live agents are running.

SECTION 04

Calculators

Every calculator below uses a formula stated in the source. Inputs are yours; results are computed in the browser and shown with the formula used.

1. Cash offer

Source formula, AutonomousCloser.sendOffer: offer = (ARV x 0.70) minus repairs, repairs assumed at 10 percent of ARV. So at the default, offer = 60 percent of ARV.

After repair value (ARV), $ Repairs, percent of ARV

Repairs: $30,000
Offer: $180,000
Offer as share of ARV: 60.00%

Hand check: 300,000 x 0.70 = 210,000; 210,000 minus 30,000 = 180,000.

2. Hunt economics

Source: cost about $1.50 per target across 4 channels; 1 in 50 converts to a $10,000 assignment fee; $10,000 / $75 = 133x ROI. Hunt every 5 minutes, 288 cycles per day, 10 targets per cycle = 2,880 contacts per day.

Targets per cycle Cycles per day Cost per target, $ Conversion: 1 in Assignment fee per conversion, $

Contacts per day: 2,880
Daily cost: $4,320.00
Conversions per day: 57.60
Revenue per day: $576,000.00
ROI multiple: 133.33x

The cost breakdown in the source sums to $1.5075, not $1.50. See the honest gaps section. At $1.5075 the same inputs give $4,341.60 cost and 132.67x.

3. Distress scores

Foreclosure: min(10, 5 + (30 minus days to auction) / 3). Code violation: min(10, 6 + total fines / 1,000).

Days to auction

Foreclosure distress: 8.00

Total code fines, $

Code violation distress: 8.50

Equity, percent Distress score, 0 to 10

Passes the v3.0 filter (distress at least 7 and equity at least 30)? YES

Hand checks: days 21 gives 5 + 9/3 = 8. Fines $2,500 gives 6 + 2.5 = 8.5. Days 0 gives min(10, 15) = 10. Days 30 gives 5.

4. Deal DAO token price and tax sale offer

Deal DAO: price per token = (acquisition cost x 1.20) / 1,000 tokens. Auto-close at 70 percent subscribed in 48 hours. Tax sale pre-emption: offer = (market value x 0.60) minus payoff, where payoff is taxes owed plus penalties.

Acquisition cost, $

Price per token: $300.00
Total raise at full subscription: $300,000.00
70 percent subscription threshold: $210,000.00

Market value, $ Tax payoff (taxes plus penalties), $

Tax sale pre-emption offer: $168,000.00

Hand checks: 250,000 x 1.2 / 1,000 = 300. 300,000 x 0.6 minus 12,000 = 168,000.

Figure 4. Cost per target by channel, from the v3.0 economics section. Bars are proportional at 400 px per dollar. The six lines sum to $1.5075.

SECTION 05

Multi-modal assault

Four channels, executed in parallel in the source code. Select a channel to inspect it. The persona name, providers, and limits below are the ones named in the source.

Select a channel

Channel	Provider in source	Cost in source	Return value in code	Missing input behavior
Email	OpenAI GPT-4 composition, Resend send	$0.02 composition, $0.00 send on free tier	SENT	No email address is not handled in the excerpt
SMS	Twilio	$0.0075 per SMS	SENT, or NO_PHONE	Returns NO_PHONE when no phone is present
Voice	Bland.ai, voice mason, 5 minute maximum	$0.09 per minute, average 2 minutes = $0.18	CALLING, or NO_PHONE	Returns NO_PHONE when no phone is present
Physical mail	Lob.com, color, single sided	$1.25 per letter	MAILED, or NO_ADDRESS	Returns NO_ADDRESS when no mailing address is present

The email sender in the source is shown as Alex at alex@neuruh-re.com, described as a rotating persona, with replies to acquisitions@neuruh-re.com. The voice script forbids being pushy, asking for personal financial information, and making specific offer amounts on the call.

SECTION 06

Capital formation and tokenization

The Capital Legion and the Tokenization Legion, as specified in v2.0. All figures are source claims.

Capital Formation Engine

Deal DAO: 1,000 tokens, 20 percent markup on acquisition cost, minimum $500, maximum $50,000 per investor, projected IRR 22 percent, 18 month hold, BRRRR strategy.
Hard money arbitrage: query 50 lenders at once, 75 percent LTV, 12 month term, maximum 12 percent rate, refinance scheduled at 6 months.
Family offices: auto-generated memo, 100 family offices by LinkedIn and email, 48 hour response watch. The memo track record in the source is stated as 47 deals and 34 percent average IRR. That is a source claim, not a verified fact.
DeFi: property NFT as collateral, 60 percent LTV, USDC, 8 percent interest against 12 percent hard money, described in the source as a perpetual term with an infinite leverage loop. That last phrase is the source wording, not a verified outcome.

Fractional Sovereignty

Token standard: ERC-3643 security token, liquid democracy governance.
Revenue split: operations 10 percent, maintenance 15 percent, investors 75 percent.
Smart property: August locks, Nest thermostat, Ring cameras, plus leak, occupancy, energy, and climate sensors.
Tenant bot: auto screening, digital leasing, AI maintenance triage, automated eviction filing if rent is more than 5 days late, as written in the source.
Secondary market: automated market maker, 2 percent fee per trade, 5 percent spread, arbitrage check every 60 seconds.

Revenue streams

Six streams from the pitch deck, with the deck share for each. They sum to 100 percent.

Stream	Deck share	Source description
Deal arbitrage	35%	Acquire at 60 to 70 percent of ARV through distress detection
Capital formation	25%	Automated syndication via Deal DAOs, 2 percent acquisition fee plus 20 percent carry
Tokenization fees	15%	ERC-3643 issuance, 3 percent minting plus ongoing management
Market intelligence	10%	Sell proprietary data feeds to institutional investors
SaaS licensing	10%	White-label swarm technology to other operators
Development	5%	Ground-up construction and value-add repositioning

The v2.0 economics section gives a different fee set: acquisition fees 3 percent, asset management 2 percent annually, disposition 20 percent of profits, tokenization 5 percent of fractional sales, secondary market 2 percent per trade, data sales $500 per lead. Both sets are reproduced; they are not reconciled in the source.

SECTION 07

Dominion economics and evolution

Worked examples using only numbers stated in the source.

Acquisition fee

$1,500,000

3 percent of $50,000,000 volume, as stated in the source.

Asset management

$2,000,000 / yr

2 percent annually on $100,000,000 AUM, as stated in the source.

Daily hunt volume

2,880 contacts

288 cycles per day x 10 targets per cycle. Note the gap: the code allows 10 per colony per cycle across 5 colonies, which would be 14,400, not 2,880.

Evolution rules

Rule	v3.0 colony	v2.0 meta legion
Terminate	Revenue under $1,000 and kills under 5	ROI under 1.0, bottom 20 percent
Breed / clone	Revenue over $10,000 spawns a new colony of the same vertical, generation plus 1, 10 percent mutation	ROI over 3.0, each strong agent spawns 3 offspring at 10 percent mutation
Clock	Colony lifespan 24 hours	Meta optimization loop every hour; market prediction horizon 90 days
Predictive gate	Not stated in v3.0	Pre-position only when confidence over 0.8 and growth over 0.15

Rent optimization, as specified

100 percent occupied for 3 months: raise rent 5 percent. Under 90 percent occupied: lower rent 3 percent and boost marketing. These are source rules, not tested outcomes.

Source claim, not a verified fact: the v2.0 endgame states that competitors pay for data, wholesalers sell to the Dominion, and tenants pay the Dominion, and that the system becomes the Amazon of real estate. No evidence for that outcome is in either source.

SECTION 08

Pitch deck explorer

The 12-slide SREIC deck inside the zip, SREIC_PitchDeck.pptx.html. All figures are the deck figures, labeled as source claims.

Cover: SREIC, 5,000 autonomous agents, one Dominion, complete market domination. Confidential, do not distribute.
Problem: $3.8T annual waste, 72 hour average response, $12,000 average customer acquisition cost, 94 percent of distressed opportunities never captured.
Solution: 5,000 plus agents, 133x ROI multiple, under 2 minute response, 24/7 operation.
Architecture: 7 legions, 5,150 agents in total. Note the cover says 5,000 while the architecture slide totals 5,150.
Market: TAM $42T, SAM $2.4T, SOM $180B.
Business model: 6 revenue streams, shares in section 06 above.
Moats: self-evolving intelligence, speed asymmetry, data network effects, counter-intelligence.
Traction roadmap: Now, MVP complete, 500 agents, 3 markets. Q2 2026, Series A close, 2,000 agents, 15 markets. Q4 2026, full legion deploy, 5,000 agents, 50 markets. 2027, tokenization live, Deal DAOs, $100M AUM. 2028, Series B or IPO, global expansion, $1B plus AUM.
Financials: Year 1, 250 deals, $12.5M gross revenue, $3.1M net profit, $25M AUM, 2,000 agents. Year 2, 1,200 deals, $68M, $20.4M, $180M, 5,000 agents. Year 3, 5,000 deals, $320M, $112M, $1.2B, 12,000 agents. Unit economics: CAC $850, LTV $113,050, LTV to CAC 133 to 1, payback under 30 days, gross margin 68 percent.
Team: Sovereign Architect, Swarm Commander, Capital Oracle, CFO background stated as ex-Goldman and ex-Blackstone, Acquisition Warlord, plus 12 engineers, 8 data scientists, and 6 real estate operators. Roles only; no personal names are in the deck.
The Ask: $10M Series A, 20 percent equity, $40M pre-money valuation, $500,000 minimum check. Use of funds: engineering 40 percent, market expansion 25 percent, data infrastructure 15 percent, tokenization 10 percent, operations and legal 10 percent.
Closing: The future is agentic, own it. Contact shown as sovereign@sreic.dom and sreic.dom/deck.

Source claim, not a verified fact: every financial projection, market size, traction date, and team background above comes from the deck file. None of it is verified by an external record in either source.

SECTION 09

Tool arsenal

The ten platforms listed at the end of the docx, in the final Kimi compilation. The earlier list in the same document counts twelve; the final compilation counts ten. Both counts are recorded in the gaps section.

#	Platform	What the source says it does
1	SREIC v2.0, core swarm	DominionCore, 7 legions, 5,150 agents, cross-legion intelligence sharing
2	ApexPredator strategy engine	Ghost offers, distress amplifier, wholesaler intercept, pre-probate, tax sale pre-emption, pre-REO capture
3	AutonomousCapitalEngine	Deal DAOs, hard money arbitrage, family office pitching, DeFi collateralization, automated capital calls
4	FractionalSovereignty	ERC-3643 minting, IoT property management, secondary markets, yield distribution, KYC and AML compliance
5	MarketWarfare Suite	SEO dominance, data poisoning, talent poaching, deal sniping, review manipulation, ad auction dominance
6	MultiModalAssault	Email, SMS, voice AI, physical mail, social bots, retargeting
7	Predictive Intelligence Engine	Pre-funding detection, layoff prediction, talent arbitrage, 50 plus distress sources, price prediction, market timing
8	SwarmEvolution	Agent breeding, mutation, performance analysis, strategy evolution, death mechanism
9	Interactive dashboard	US map, 3D legion icons, live feed, performance chart, spawn controls, command console, system logs
10	VC pitch deck	12 slides, financial projections, market analysis, team, $10M Series A ask

Adversarial items in this list, such as data poisoning, review manipulation, and reputation warfare, are reproduced as source content for documentation. They are not instructions on this page, and no tool on this page performs them. Several would raise legal, platform, and compliance questions if built as described; the source does not address those questions.

SECTION 10

Try it walkthrough

Six steps, about two minutes, all inside this page.

Load Sample A below. It selects the Acquisition Legion, the probate vertical, Phoenix, and an ARV of $300,000.
Read the offer calculator. It should show $180,000, which is 60 percent of ARV at the default 10 percent repairs.
Change days to auction to 0 in the distress calculator. Foreclosure distress should read 10.00. Change it to 30. It should read 5.00.
Select a legion row, then press Spawn +50. The legion count and the design total change on this page only.
Click a city dot on the map, then run a simulated hunt cycle and read the feed.
Press Copy share link. The link encodes your inputs in the URL hash and restores them on load.

Sample loaders

No sample loaded yet.

Completeness meter

Fill the eight fields the model needs for a complete simulated hunt plan: legion, vertical, city, ARV, equity, distress score, channel, and acquisition cost.

0% complete, 0 of 8 fields

SECTION 12

How it works, and what this is not

Plain reading of the model, with the boundary stated.

How it works

Pick a legion and a vertical. The legion sets who acts; the vertical sets what signal is hunted.
Enter a property value, equity, and distress. The filter passes only distress at least 7 with equity at least 30 percent.
Compute the offer, the hunt cost, and the token price with the source formulas in section 04.
Run a simulated hunt cycle. The feed writes deterministic lines from your selections. Nothing is sent.
Copy the share link to save or send the exact inputs.

What this is not

Not a live swarm. 5,150 is a historical logical-agent design target, not 5,150 live sessions.
Not connected to PropStream, RealtyTrac, Trellis, OpenAI, Resend, Twilio, Bland, Lob, PandaDoc, or DocuSign. No API keys are in this page.
Not a real offer, contract, or outreach. No email, text, call, or letter is sent.
Not verified performance. ROI, revenue, deal counts, and market sizes are source claims.
Not legal or compliance advice. Several source tactics would need legal review before any real use.

Related Neuruh surfaces: Neuruh Hub, Real Estate Agent hub. The hub is the master map for the Neuruh builds.

SECTION 13

Honest gaps

Discrepancies and missing pieces found while reading both sources in full. Nothing here was silently reconciled.

5,000 versus 5,150: the deck cover and the launch script say 5,000 agents; the legion table and the architecture slide total 5,150. This page uses 5,150 as the design target and notes the 5,000 figure where it appears.
Acquisition sub-units do not sum: the Dominion interface lists probate hunters 500, foreclosure sharks 300, distress signals 200, and off-market snipers 1,000, which sum to 2,000, not the 2,550 spawn count for the Acquisition Legion. 550 agents are unaccounted for in the source.
Cost per target: the channel costs sum to $1.5075, while the economics section rounds to $1.50. ROI at $1.5075 is 132.67x, not the stated 133x. The calculator defaults to $1.50 to reproduce the source figure and shows the $1.5075 result in the note.
Contacts per day: 10 targets per cycle x 288 cycles = 2,880, but the code allows 10 targets per colony per cycle across 5 colonies, which would be 14,400 contacts per day. The source does not say which reading is intended.
Offer example mismatch: the closer stub returns a $200,000 offer for a $300,000 ARV property, while its own formula gives $180,000. The formula is used here.
Tool count: one compilation in the docx lists 12 platforms and 50 plus tools; the final compilation lists 10 platforms and 40 plus tools. This page follows the final list of 10.
Dashboard numbers are seeded, not observed: App.tsx starts with 1,247 deals and $48,750,000 revenue and then adds random increments every 2 seconds. The map places 100 agents at random among 48 cities. None of that is a measured result.
Missing in the sources: no API keys or live endpoints, no verified deal records, no county or court data, no compliance review for the adversarial tactics, no D1 schema for the Dominion sharding beyond the named databases, and no test or deployment receipt for the swarm itself.
Images not reproduced: the zip contains five deck images (cover, architecture, problem and solution, traction, vision). This page uses no images, per the build standard, so those visuals are described, not shown.
Layout not eyeballed: this page was tested in jsdom for logic and controls, not viewed in a real browser. Layout is an accepted limit.

Source files: Agent_Swarm_Domination__1.docx and Kimi_Agent_Agent_Swarm_Domination.zip, both read in full on 2026-11-04. Built as a standalone demo at agent-swarm-domination.pages.dev. See the Neuruh Hub for the full build map.