MythadisSwarm

Private R&D

Private R&D

Current State

Mythadis separates demonstrated capability from active development, research direction, and aspiration.

State what is true now.

Mythadis Swarm is an early private research and engineering project. The purpose of this page is to provide a clear, current record of what has been demonstrated, what is being built, and what remains unproven.

Early. Private. Measurable.

LAST REVIEWED / JULY 2026CURRENT STATE VERSION / 0.1STATUS / PRIVATE R&D
  • PRIVATE R&D
  • EARLY MULTI-NODE PROTOTYPE
  • NO PUBLIC AGENT
  • NOT PRODUCTION-READY

Mythadis Swarm currently operates as a private research prototype across a small number of heterogeneous machines.

Current work is focused on proving the underlying coordination fabric before expanding into public participation or mission-scale systems.

What has been demonstrated

Peer Discovery

DEMONSTRATED PROTOTYPE
Scope
Multiple nodes have discovered and recognized one another within private test environments.
Evidence type
Private Mythadis Swarm repository documents local and Docker Compose node operation with bootstrap configuration and libp2p communication.
Limitation
Validated in controlled private environments, not a public network.
Next Gate
Publish a bounded Field Report with test conditions and observed node behavior.

Cross-Platform Operation

DEMONSTRATED PROTOTYPE
Scope
Early node software has operated across Linux and macOS systems using different hardware profiles.
Evidence type
The prototype supports local or containerized node operation and has been used across heterogeneous private systems.
Limitation
This is not a supported public installer or production distribution.
Next Gate
Record reproducible setup notes for supported private test profiles.

Multi-Node Connectivity

DEMONSTRATED PROTOTYPE
Scope
Private tests have established communication among multiple nodes, including nodes operating across different local subnets.
Evidence type
Swarm demos define multiple node services and control endpoints for seed, engineer, scientist, mythadis, and longevity profiles.
Limitation
Connectivity claims are bounded to private trials and demo environments.
Next Gate
Validate longer-running multi-node communication with preserved logs.

Distributed Data Experiments

EXPERIMENTAL
Scope
Early experiments have explored distributed storage, data fragmentation, recovery, and retrieval across participating nodes.
Evidence type
The private prototype uses Kademlia storage experiments and Reed-Solomon sharding over Data Parcels.
Limitation
Distributed storage is not described as permanent, globally durable, or production-secure.
Next Gate
Demonstrate recovery under node loss with defined shard conditions.

Local Persistence

DEMONSTRATED PROTOTYPE
Scope
Node identity and local data stores persist across local prototype runs.
Evidence type
The private repository documents node identity storage and sled data under the configured data directory.
Limitation
Persistence is local prototype behavior, not a backup, compliance, or durability guarantee.
Next Gate
Document recovery behavior across restart, corruption, and data-directory migration scenarios.

Lightweight Knowledge Discovery

EXPERIMENTAL
Scope
Early query paths can match ingested evidence and return source-grounded responses when semantic context exists.
Evidence type
The private prototype documents hunt, ingest, ask, catalog, parcel lookup, content hash, score, source preview, and matched/unmatched response shapes.
Limitation
This is not a generalized research engine, citation system, or truth engine.
Next Gate
Publish a field report showing source-grounded response behavior against fixed demo data.

What is being built now

IN ACTIVE DEVELOPMENT

Data Parcels

A self-describing, content-addressed data model intended to preserve identity, context, provenance, relationships, and integrity.

IN ACTIVE DEVELOPMENT

Archive Health

A system for distinguishing healthy, degraded, at-risk, and unrecoverable archive states.

IN ACTIVE DEVELOPMENT

Mythadis Agent Foundations

Early work toward a participant-facing Agent with explicit local resource and policy boundaries.

IN ACTIVE DEVELOPMENT

Mission Control Foundations

Initial orchestration capabilities for agents, health, roles, missions, storage, compute, and validation.

IN ACTIVE DEVELOPMENT

Distributed Data Path

Continued validation of storage, transfer, verification, and recovery across multiple nodes.

What remains a research direction

Swarm Compute at meaningful scale

Coordinated compute across many participants remains unclaimed.

RESEARCH DIRECTION

Must prove: Scheduling, workload isolation, verification, recovery, security, and measurable efficiency.

Concordance Engine

A future system for reasoning over agreement, disagreement, evidence, and unresolved claims.

RESEARCH DIRECTION

Must prove: Reliable evidence handling, dissent preservation, traceable outputs, and validation across real use cases.

Aegis

A future safety, policy, and trust boundary for coordinated intelligence.

RESEARCH DIRECTION

Must prove: Threat modeling, enforceable policy boundaries, auditability, and failure behavior.

Public Agent distribution

The Mythadis Agent is not publicly available.

FUTURE GATE

Must prove: Local safety, installation, update behavior, privacy boundaries, and clear operator control.

Mission Swarms

Mission Swarms remain future work, not demonstrated public outcomes.

FUTURE GATE

Must prove: Mission definition, participant roles, evidence flows, coordination loops, and measurable outputs.

Large-scale archive preservation

The current prototype does not claim permanent archive preservation.

RESEARCH DIRECTION

Must prove: Durability, recovery, integrity, redundancy, governance, and long-running operation.

Cross-institution coordination

Coordination across institutions remains a future research and partnership challenge.

RESEARCH DIRECTION

Must prove: Governance, permissions, data boundaries, trust model, and repeatable collaboration workflows.

Public volunteer network

No public volunteer network currently exists.

FUTURE GATE

Must prove: Safety, onboarding, incentives or motivation model, support, governance, and abuse resistance.

What we are not claiming

  • Mythadis Swarm is not production-ready.
  • No public volunteer network currently exists.
  • The Mythadis Agent is not publicly available.
  • Mythadis does not currently operate at global scale.
  • Mythadis has not demonstrated humanity-scale mission outcomes.
  • Mythadis does not currently solve longevity, climate, pollution, medicine, or public science.
  • AI agreement is not treated as proof.
  • Distributed storage is not described as permanent.
  • Decentralized architecture is not presented as automatically secure or trustworthy.
  • No investment return or token model is being promoted.

The purpose of these boundaries is not to make the vision smaller. It is to make future proof meaningful.

Current private trials use heterogeneous systems.

Current private trials use a small collection of heterogeneous systems, including Linux and macOS nodes with different processor, memory, storage, and network characteristics.

These environments are used to test discovery, connectivity, stability, data movement, recovery, and increasingly coordinated work.

PRIVATE TEST ENVIRONMENT / CONCEPTUAL VIEW

How capability becomes a claim

  1. Define the bounded capability.
  2. Declare the test conditions.
  3. Run the test.
  4. Preserve logs, outputs, and artifacts.
  5. Record limitations and dissent.
  6. Publish an appropriate Proof entry or Relic.
  7. Update Current State only after validation.

A capability enters this page only when there is evidence that it exists under defined conditions.

What is being proven next

Three-node distributed data-path validationLong-running node stabilityRecovery under node lossAgent role clarityMission Control observabilityData Parcel contractArchive Health measurementReproducible deployment

Each priority must produce bounded evidence before it advances the public claim.

A living record

Current State is intended to change as evidence changes.

Capabilities may advance, remain experimental, become contested, or be removed when later evidence shows that an earlier interpretation was incomplete.

Help test what comes next.

Mythadis is interested in engineers, researchers, systems thinkers, and mission partners who value ambitious goals, explicit limitations, and evidence-based progress.