SELECTED WORK / LIVING MEMORY
Context should not disappear.
A local-first foundation for verifiable technical memory: it preserves what a project knows about itself while keeping its sources, limits, and uncertainty visible.
01 / CONTINUITY
The problem
Projects change faster than the memory of the people building them. Decisions, sources, and conditions become scattered; reconstructing why something is true becomes costly or impossible.
02 / MEMORY LAYER
Persistent technical memory
Living Memory defines a layer to preserve and verify claims, sources, provenance, uncertainty, contradictions, and change over time. Evidence comes before narrative.
03 / PROVENANCE
A claim with provenance
Provenance connects a claim, evidence, location, source, rule, intermediate artifact, snapshot, version, and review. The model makes explicit what supports something and what remains uninspected.
04 / REPRODUCIBILITY
Determinism before interpretation
The validated capability uses a strict project profile, a confined and authorized inspection root, a deterministic inventory, SHA-256 integrity, and canonical JSON scan inventory and manifest output. Broader analysis remains out of scope.
05 / BOUNDARY
Public and private are not the same
The public surface shares purpose, principles, validated scope, high-level architecture, selected evidence, and historical milestones. Current source code, detailed operational documentation, security-sensitive mechanisms, internal project memory, and selected implementation details remain private or proprietary.
06 / STATUS
Public status
Living Memory is in active product development. The current foundation implements explicitly validated boundaries without presenting itself as a complete application or production-ready service.
07 / PUBLIC EVIDENCE
Available public evidence
The authorized public repository documents Living Memory's purpose, principles, validated scope, publication boundaries, and historical evidence.
LIVING MEMORY / REPOSITORY ↗