Reference ecosystem · Evidence snapshot

A protocol extracted from systems that run.

Kynetra implements the initial Web7 reference stack. HyperBridge applies that stack to production platforms. Neither company defines the limits of the open protocol.

Published evidence · July 2026

A larger proving ground than the old Web7 map showed.

The previous ecosystem page mixed applications, frameworks, protocol layers, and research targets. The revised map reports each organization and product according to the job it actually performs.
39 applications in the published Kynetra ecosystem Kynetra ecosystem ↗
24 Kynetra applications reported live today Live directory ↗
40+ HyperBridge production platforms HyperBridge Digital ↗
6 Kynetra product domains represented publicly Platform map ↗
Recent advances

What changed around Web7.

The ecosystem now covers the full accountability lifecycle. That makes a narrower, stronger protocol position possible.
01 Governed control plane Prime now centralizes planning, tenant boundaries, policy, approvals, evidence gates, and controlled learning. Prime
02 Portable evidence plane KynetraDB and Search bring events, memory, vectors, graphs, lineage, hybrid retrieval, and proof records behind versioned contracts. DB + Search
03 Model lifecycle KynML, Foundry, and Deploy separate training intent, evaluation, promotion, release execution, and rollback. ML lifecycle
04 Bounded action runtime Kynetra FX, KYRx, and HBForge turn approved plans into typed APIs, portable handlers, and auditable modules. Action plane
05 Operator control AIStudio, Resolve, Browser, and Pods expose plans, approvals, audited web actions, incidents, and workload constraints. Operations
06 Production pressure QuantumOS and the wider HyperBridge portfolio exercise identity, commerce, finance, operations, content, and compliance workflows. Proof
Reference directory

One job per product.

Status describes the product surface today. A live website does not automatically mean every advertised capability is production-complete.
Kynetra Prime ↗ Plans, Regents, policy, approvals, accountability events, and governed learning. control plane active
AIStudio ↗ Local-first model workstation and operator command center for agents, code, retrieval, and approvals. desktop + web active
Kynetra Auth ↗ Identity, sessions, tenant claims, roles, delegation, service trust, and audit. identity boundary live
KTL ↗ Consent, policy, compliance evidence, approval quorums, and tamper-evident trust records. governance layer live
KynetraDB ↗ Portable documents, events, memory, vectors, graph, files, lineage, and proof records. data + evidence development
Kynetra Search ↗ Tenant-isolated lexical and semantic retrieval, connectors, ranking, feedback, and answer traces. retrieval plane live
KynML ↗ Compiler-first training specifications with validation, deterministic PyTorch generation, formatting, and LSP support. training intent executable
Kynetra Foundry ↗ Experiment signals, evaluation evidence, bounded improvement proposals, and model promotion candidates. evaluation gate development
Kynetra Deploy Cloud jobs, release plans, provider execution, canaries, rollback, and runtime operations. release execution development
Kynetra FX Contract-first edge backend, typed routes, OpenAPI, clients, capability ports, tenant isolation, and hybrid RAG. action gateway development
KYRx ↗ Language and compiler contracts for typed workflows, policies, native interop, and deployable artifacts. compiler boundary active
HBForge ↗ Seventy-seven owned application modules with zero npm dependencies, used across HyperBridge platforms. 3 years in production production
Kynetra Resolve ↗ Incident and request intake, deduplication, triage, SLA ownership, root cause, and governed remediation. recovery plane live
Kynetra Browser ↗ Permission-brokered browser automation, structured extraction, monitoring, isolated profiles, and hash-chained audit. phase 1 private beta
Kynetra Pods Rootless workload packaging, OCI plans, database constraints, native desktop, and provider-aware validation. early implementation development
QuantumOS X3 ↗ Commerce and operations system exercising tenant workflows, orders, inventory, storefronts, finance, and growth. production proving ground production
Independence

The reference stack is replaceable by design.

A conforming implementation may use a different identity provider, model runtime, application framework, database, search engine, cloud, or settlement rail. Interoperability comes from the receipts—not brand alignment.
Adopt

Use the complete stack

Start with Kynetra’s reference products when their operating boundaries match your requirements.

Integrate

Wrap an existing system

Add Web7 authority, policy, trace, and outcome receipts around actions your current platform already performs.

Implement

Build another conforming stack

Use the RFCs and forthcoming fixtures to create an independent implementation without Kynetra dependencies.

Choose your path

Use the reference stack—or prove yours interoperates.

The Foundation’s job is to make both routes possible and independently testable.