Mycelix

Infrastructure for agent-centric coordination

Mycelix explores decentralized identity, consent, reputation, commons management, and collective decision-making on Holochain. Some infrastructure is implemented and tested; its governance mechanisms remain experimental and unvalidated for real-world use.

Core data structures
Implemented

Unit testing
Passing at reviewed commit

Multi-agent conductor testing
Partial

Cross-cluster integration
Experimental

Civic governance validity
Not established

Security review
Internal only

Independent governance deployment
None

Known critical blockers
Open — see below

A readiness snapshot, not a scale claim. Project size (zome/test/line counts) lives in the maturity matrix, not here — size doesn't establish reliability.

See what works today Review risks and blockers Inspect the source

Three different things, on one site

This project mixes working code, untested-at-scale governance mechanisms, and a normative political argument. Conflating them is how a research prototype ends up sounding like a finished claim about how societies should be run. Here's the split, and where each lives:

Status words used across this site mean specific things, not vibes:

Implemented Exercised internally Replicated internally Independently reproduced Proposed Blocked Superseded Falsified

Implemented infrastructure

Implemented code with unit-test coverage

Holochain hApp architecture across 18 named domain and project areas (see Architecture below — this org's own README describes the unified hApp bundle as having 10 roles, and the exact taxonomy isn't yet generated from one canonical manifest), three experimental economic mechanisms implemented in code, DID-based identity, and a bridge protocol for cross-cluster calls.

Builds and has unit tests today — Implemented
A minority of clusters have real multi-agent (Tryorama/sweettest) coverage — Partial
Two areas have working user-facing application prototypes: Pulse and Praxis — see below

Experimental governance mechanisms

Designed, partly built, unproven

The participation profile and vote-weight-by-tier design, detailed on the Governance page, are being tested in code — not a validated model of legitimate governance.

Never governed a real community or made a binding decision for real people
Open, unresolved risks — full register on the Risks page
Presented for scrutiny, not as a finished answer

Normative philosophical proposal

An argument, not a finding

The Philosophy page holds this project's own political and philosophical position — offered for debate, not as evidence.

Not empirically established, and not claimed to be
You're meant to disagree with parts of it if you do
Kept on its own page so agreeing with the code doesn't require agreeing with the politics

Architecture, in brief

Four layers, increasing in maturity from top to bottom. Solid = implemented, dashed = partial, dotted = proposed-only.

Applications                                    (▫▫▫ proposed / ▬▬ partial)
  Pulse (live email) · Praxis (live curriculum) · Marketplace, Commons UIs (mostly unbuilt)
                          │
Domain clusters                                 (━━ implemented, unit-tested)
  Identity · Reputation · Knowledge · Finance · Civic · Health · … (18 total)
                          │
Shared protocol layer                           (━━ implemented / ▬▬ partial coverage)
  Authorization · Evidence · Consent · Cross-cluster bridge contracts
                          │
Holochain                                       (━━ implemented — upstream)
  Source chains · DHT · Agent-centric validation

Taxonomy count: 18 named domain and project areas appear in this repository today (Identity, Governance, Commons, Civic, Hearth, Personal, Finance, Attribution, Marketplace, Supply Chain, Knowledge, Health, EduNet, DeSci, Energy, Climate, Space, Music) — not all of these are necessarily "Holochain clusters" in the strict architectural sense; the org's own README separately describes a 10-role unified hApp bundle. Earlier org materials, including parts of this site, have said "16," a number that predates several of these areas. Until the taxonomy is generated directly from canonical manifests rather than hand-counted, treat both "16" and "18" as approximate. Full per-cluster status: maturity matrix.

Three blockers to any real deployment

The short version — full register with severity, owner, and evidence links is on the Risks page.

No appeals or contestability process for participation-profile decisions. Blocks any deployment that affects real people.
Guardian emergency authority is unimplemented and governance-blocked — no approved scope, quorum, duration, audit, or revocation mechanism exists.
A known authorization-binding defect class is still open in some clusters (fixed in most, tracked cluster-by-cluster) — see Security status.

The Ecosystem

Luminous Dynamics

The parent research studio.

Symthaea

Cognitive architecture research. HDC + IIT-inspired measurement + CfC dynamics + active inference — experimental, claim-disciplined.

Eight Harmonies

The philosophical foundation this project draws on, kept separate from the code's claims.

Nix-adjacent work

nix-signature-policy is this org's active Nix-adjacent project today (the earlier Luminous Nix is archived).