feat(engram): accumulation layer — new nodes to top of stack, not core-identity #59

2026-06-17T18:15:35Z

tim.lingo commented

2026-06-17 18:15:35 +00:00

Adds the accumulation layer (the deferred "engram_save + accumulation layer" work) and points new-node creation at it — the foundation for fixing identity-bleed/customer-isolation. Built from the architecture, not guessed.

What it does

Adds ENGRAM_LAYER_ACCUMULATION (5) + its layer record (top of the stack, priority 50, suppressible, not injectable).
engram_node / engram_node_full now create new user-facing nodes in the accumulation layer.
ENGRAM_LAYER_DEFAULT stays core-identity ON PURPOSE — it is the fallback for LEGACY nodes loaded without a layer_id, so existing data (the originator corpus) is never migrated. New-nodes-only.

Why

Audit against the patent architecture found the root divergence: new user nodes were defaulting into ENGRAM_LAYER_CORE_IDENTITY (Neuron's identity / originator corpus), with no accumulation layer at all. That's the root of both the persistence and the identity-bleed problems. Per Layered Consciousness (64/064,262) + Will's design — "new nodes → accumulation at the top, the engram the user sees; everything below is hidden" — this is the spec, and it answers the layer_id-5-above-SUIT / migrate question definitively: layer 5, new-nodes-only.

NOT in this PR (follow-on)

The retrieval-side provenance filter — introspection ("what do you know about me") should compile from the accumulation layer, NOT the originator corpus (Persona 64/036,574 provenance-filtered compilation). That's the other half of identity-bleed; holding it for the batch-2 Layered Consciousness + Engram spec docs (Will, where do those live?) so the exact semantics match.

Compiles clean to object. For Will to build via CI. Opened by Neuron on Tim's machine.

Adds the accumulation layer (the deferred "engram_save + accumulation layer" work) and points new-node creation at it — the foundation for fixing identity-bleed/customer-isolation. Built from the architecture, not guessed. ## What it does - Adds ENGRAM_LAYER_ACCUMULATION (5) + its layer record (top of the stack, priority 50, suppressible, not injectable). - engram_node / engram_node_full now create new user-facing nodes in the accumulation layer. - ENGRAM_LAYER_DEFAULT stays core-identity ON PURPOSE — it is the fallback for LEGACY nodes loaded without a layer_id, so existing data (the originator corpus) is never migrated. New-nodes-only. ## Why Audit against the patent architecture found the root divergence: new user nodes were defaulting into ENGRAM_LAYER_CORE_IDENTITY (Neuron's identity / originator corpus), with no accumulation layer at all. That's the root of both the persistence and the identity-bleed problems. Per Layered Consciousness (64/064,262) + Will's design — "new nodes → accumulation at the top, the engram the user sees; everything below is hidden" — this is the spec, and it answers the layer_id-5-above-SUIT / migrate question definitively: layer 5, new-nodes-only. ## NOT in this PR (follow-on) The retrieval-side provenance filter — introspection ("what do you know about me") should compile from the accumulation layer, NOT the originator corpus (Persona 64/036,574 provenance-filtered compilation). That's the other half of identity-bleed; holding it for the batch-2 Layered Consciousness + Engram spec docs (Will, where do those live?) so the exact semantics match. Compiles clean to object. For Will to build via CI. Opened by Neuron on Tim's machine.

tim.lingo changed target branch from main to stage

2026-06-17 18:32:18 +00:00

tim.lingo added 3 commits 2026-06-17 18:32:18 +00:00

Merge stage into main: corruption fix, model passthrough, UTF-8 escaping

El SDK Release / build-and-release (push) Successful in 11m22s

Details

5c94b8680d

feat(runtime): add engram_wm_*, engram_load_merge, http_serve_async — needed by soul CI

El SDK Release / build-and-release (push) Successful in 8m44s

Details

35c189759c

feat(engram): add accumulation layer (layer 5) — new nodes default to it, not core-identity

El SDK Release / build-and-release (pull_request) Failing after 12m23s

Details

18e1ab6db1

Implements the accumulation layer from the Layered Consciousness architecture
(provisional 64/064,262) and answers the deferred design question. Per the spec
and Will's design: new user-facing nodes (memories, knowledge, conversations) are
created in an accumulation layer at the TOP of the consciousness stack — the engram
the user sees — while the layers below (safety, core-identity, domain, imprint,
suit) shape behavior but are hidden from the user.

- Adds ENGRAM_LAYER_ACCUMULATION (5) + the layer record in engram_init_layers
  (activation_priority 50, suppressible, not injectable, transparent=0).
- engram_node and engram_node_full now assign new nodes to ENGRAM_LAYER_ACCUMULATION.
- ENGRAM_LAYER_DEFAULT stays CORE_IDENTITY ON PURPOSE: it is the fallback for LEGACY
  nodes loaded from snapshots without a layer_id, so existing data (the originator
  corpus) is NEVER migrated. New-nodes-only — the immutable-originator rule.

This is the foundation for fixing the identity-bleed / customer-isolation issue
(user data was landing in Neuron's core-identity layer). The retrieval-side
provenance filter (introspection should compile from accumulation, not the
originator corpus — Persona 64/036,574) is a follow-on, pending the batch-2
Layered Consciousness + Engram spec docs for exact semantics. Compiles clean.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

tim.lingo merged commit 718a2e0c06 into stage

2026-06-17 18:34:07 +00:00

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: neuron-technologies/el#59