FLUX WIKI

Protocol documentation for the FLUX photographic archive system.

[↓ COMPLETE MARKDOWN][↓ COMPLETE PDF]

FOUNDATION

Layer 1 — What FLUX is and where it comes from

PageDescription
WHAT IS FLUXDefinition, core sentence, and philosophical foundation of FLUX.
MANIFESTOPhilosophy, lineage, and institutional statement of FLUX.
INFLUENCESHistorical and philosophical lineage of FLUX — Heraclitus, Atget, Provoke, On Kawara, Moriyama, Becher.

PROTOCOL

Layer 2 — The method and the format it produces

PageDescription
PROTOCOLThe FLUX method — steps, principles, and what the protocol enforces.
ZINE SPECIFICATIONPhysical format, page order, fixed image box rules, contact sheet, and print assembly.
ARCHIVEDigital archive structure, URL routing, timeline logic, and what is downloadable.

FIELD

Layer 3 — Where the protocol meets the city

PageDescription
FIELD ASSIGNMENTSConstraint-based documentary exercises for FLUX photographers — 36 location, method, collaborative, and directed protocols for Philadelphia and beyond.
PROJECTSCollaborative and site-specific FLUX projects — Broad Street In Flux as the founding prototype.
BROAD STREET IN FLUXCase study — the first canonical collaborative FLUX project. Philadelphia, May 10, 2026.
MARKET STREET — DIGITAL MIRRORProject assignment — a chronological documentary walk through Market Street, Philadelphia. Photography, metadata, platform data, and the city as machine-readable system.
DISPATCHThe FLUX Dispatch — 24-frame thermal accordion field document, the manila folder archive structure, ephemera protocol, and the layered document system.
GALLERY PROTOCOLSingle-image archival exhibition objects — the ephemeral print and the permanent archive as simultaneous outputs.
FIELD NODESFLUX Field Nodes — physical QR markers embedded in the city that turn street corners into open photographic archive points.

INFRASTRUCTURE

Layer 4 — The systems that run the archive

PageDescription
GENERATORPersonal generator and public generator — how they work, how they differ, and the future public publishing system.
PUBLIC CATALOGThe public FLUX catalog — community issues, submission protocol, and the CAT_NNN namespace.
NODESFLUX Nodes — physical archive terminals connected to the live FLUX publishing system.
REMOTE ACCESSPrivate remote access to the FLUX Portal via Tailscale — iPhone on cellular, no public exposure, no open ports.
BOOTSTRAPHardware infrastructure, NAS architecture, corpus migration, keeper system, embeddings, autonomous ingest pipeline, and the long-term vision for FLUX as a self-organizing photographic archive.

INTELLIGENCE

Layer 5 — Embeddings, keepers, and the personal taste model

PageDescription
INTELLIGENCELayer overview for FLUX intelligence — links to subdocuments on embeddings, keeper model, metadata enrichment, autonomous sequencing, and training data.
EMBEDDINGSTechnical specification for the FLUX visual embeddings layer — model selection, vector storage, generation pipeline, and query interface.
KEEPER MODELSpecification for the FLUX personal taste model — what it learns, how it is trained, and what it cannot replace.
METADATA ENRICHMENTFull specification for the FLUX metadata layer — database schema, enrichment pipeline, EXIF extraction, geocoding, weather, tagging, and manifest generation.
AUTONOMOUS SEQUENCINGHow FLUX sequences issue photographs — the canonical chronological default, experimental AI alternatives, and the limits of what a machine can sequence.
TRAINING DATAThe FLUX keeper archive as a machine learning dataset — structure, matching strategy, class imbalance, near-misses, and dataset construction.

PHYSICAL

Layer 6 — The autonomous print layer

PageDescription
PHYSICALThe autonomous print layer — how FLUX materializes itself as physical objects, the printer philosophy, the return-home ritual, and the long-term vision of self-printing archives.

PRESERVATION

Layer 7 — Cryptographic verification and layered permanence

PageDescription
PRESERVATION PROTOCOLBlockchain as witness, layered permanence, and the chain of visual custody — FLUX as archival infrastructure for the AI era.

ROADMAP

Layer 8 — Where FLUX is going

PageDescription
ROADMAPOpen-source vision, future software architecture, AI selection, infrastructure, and release stages.

GOVERNANCE

Layer 9 — Protocol decisions, changelog, and version history

PageDescription
CHANGELOGChronological record of all protocol decisions, infrastructure changes, and version releases in the FLUX archive system.
DECISIONS LOGRFC-style record of canonical protocol decisions — what was decided, why, and what it constrains.
VERSIONSVersion reference for the FLUX protocol, implementation, and wiki — numbering philosophy, locked semantics, and current status.

OPEN SOURCE

Layer 10 — Node architecture and future reproducibility

PageDescription
OPEN SOURCENode architecture specification for the FLUX protocol — what a FLUX node requires, what must be built before public release, and the open source philosophy.

STARTER KIT

Layer 11 — Launch your own city-in-flux geotag archive (teachable blueprint)

PageDescription
STARTER KIT — AI IMPLEMENTATION GUIDEEntry point for an AI agent (Claude Code) recreating a CITY_IN_FLUX geotag-publishing system from this wiki as the source of truth.
STARTER KIT — DOWNLOADDownload the CITY IN FLUX starter kit and launch your own city archive with Claude Code on your own AWS account.
STARTER KIT — ARCHITECTURE & CONTRACTSComponent contracts an AI agent implements — the Worker pipeline (draft generation) and the page/hub/embed/catalog builders, with inputs, outputs, and invariants.
STARTER KIT — CONFIGURATION & CONTRACTSThe implementation contracts — environment variables, the canonical-naming algorithm, and the catalog.json + submission.json data schemas.
STARTER KIT — SETUP (AGENT STEPS)Agent-executable setup — Python, AWS CLI + least-privilege IAM, S3, CloudFront, .env, and the local portal, each with verification.
STARTER KIT — DEPLOYMENT (PUBLISH CONTRACT)The publish algorithm, S3 key layout, cache-header matrix, CloudFront invalidation contract, and the hub+embed sync invariant.
STARTER KIT — PROJECT WORKFLOW (AGENT RUN)Run one project end to end — name, upload, generate, publish — via CLI or portal, with per-step verification and the alive-vs-stuck diagnostic.
STARTER KIT — COMMON MODIFICATIONSRecipes for the changes operators ask for — each with what to change, where, the invariant to respect, and an acceptance test.
STARTER KIT — SECURITY & SANITIZATIONWhat never ships — secrets, personal paths, real catalog data, AWS keys — and how the kit is sanitized.
STARTER KIT — TEMPLATE KIT PLANWhat goes in the template — reusable code, excluded proprietary systems, sample data, .env.example, and required docs.
STARTER KIT — ROADMAPRelease path for the kit — private template zip first, then a workshop resource, then a possible open-source core.