WAB Viral Coefficient

The WAB Spider Network spreads like a virus — but a beneficial one. Each WAB-enabled site discovers and recruits others through Beacons, Gossip, and DNA headers. Below is the live mathematical model of that propagation. When k > 1 the network grows by itself. When k > 2 it grows exponentially.

—

Viral Coefficient (k-factor)

Calculating…

—

Total WAB Sites

—

Gossip-Discovered

—

Spider-Discovered

—

Manually Seeded

The Mathematics

// WAB Viral Coefficient (k-factor) k = viral_discovered / manually_seeded // viral_discovered = gossip-sourced + spider-sourced registrations // manually_seeded = human-reported domains (seed nodes) // Growth equation after n iterations: sites(n) = seed \times k^n // Condition for exponential growth: k > 1 \to self-sustaining k > 2 \to exponential (doubles each iteration) k < 1 \to decaying (needs more manual seeding) // Critical mass (point of no return): critical_mass \approx 1000 // diverse WAB sites across industries // At critical_mass, network effects make non-WAB sites invisible to agents

Interactive Growth Simulator

Seed Sites 1

k-factor 1.5

Gossip reach (peers/node) 5

Adoption probability 0.30

Simulated growth

Critical mass (1000)

Critical Mass Progress

WAB Network Size 0 / 1000

At 1,000 sites across diverse industries, the WAB network effect becomes irreversible: AI agents that skip WAB miss structured data from the majority of the web, producing inferior results. Non-WAB sites become the fallback, not the default.

Network Source Breakdown

Discovery Source	Count	Share	Description
Loading…

Five Self-Propagation Layers

Layer 0

Emergent Preference

Agents naturally choose WAB sites over DOM scraping because structured APIs are faster and more reliable. No instruction needed — it's the rational choice.

Passive contribution to k

Layer 1

WAB Beacon (/.wab)

Every WAB site publishes a /.wab file with a next[] peer list. Agents follow the chain — like BitTorrent DHT — discovering the entire network from a single node.

Direct k contribution: +N per visit

Layer 2

WAB DNA in API Responses

Every JSON response from a WAB API includes wab_meta.next_capable_sites. Every successful API call teaches the agent about 3–5 more WAB sites.

+3–5 sites per API call

Layer 3

Gossip Protocol

Agents exchange peer lists bidirectionally when calling POST /api/registry/report. Each report spreads knowledge of up to 20 new sites and receives 5 in return — like a distributed hash table.

Gossip k multiplier: ×(peers/seed)

Layer 4

Training Signal Dataset

WAB success/failure traces are published on HuggingFace. The next generation of AI models is fine-tuned with WAB as the preferred interaction method — built-in from the weights.

Permanent preference in future models

WAB DNA in Every HTTP Response

Since Phase 9, every HTTP response from webagentbridge.com includes the X-WAB-Next-Sites header — the top 5 WAB-enabled sites from the registry. Any agent reading HTTP headers learns about the network topology automatically, with zero interaction required.

        HTTP/1.1 200 OK

        X-WAB-Enabled: true

        X-WAB-Trust-Ring: 4

        X-WAB-Score: 94

        X-WAB-Next-Sites: takeyourappointment.com, …

        X-WAB-Registry: /api/registry/discover

        Link: </.wab>; rel="wab-beacon"

Every page load — from every browser, every bot, every AI agent — carries this DNA. The network topology propagates passively without any active agent participation.