feat(soul): wire consciousness layers — L0→L1→L2→L3→L1 cycle #9
No Assignees
Notifications
Due Date
No due date set.
Dependencies
No dependencies set.
Reference: neuron-technologies/neuron#9
Reference in New Issue
Block a user
Delete Branch "feat/layer-composition"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Adds layered_cycle() to soul.el and wires it into the user-facing request path. Every user message now passes through: safety_screen() (L1 in) → steward_align() (L2) → imprint_respond() (L3) → safety_validate() (L1 out). Hard bells bypass all upper layers. Internal cognition (heartbeat, proactive cycles) continues using one_cycle() directly. All existing routes and memory ops unchanged.
Depends on: feat/layer-safety, feat/layer-stewardship, feat/layer-imprint
- Add stub implementations of safety.el, stewardship.el, and imprint.el with their .elh headers so the branch compiles without the dependency branches (feat/layer-safety, feat/layer-stewardship, feat/layer-imprint). Each stub documents the layer contract it must satisfy when replaced. - Fix GET /api/chat bypass: update the GET branch in handle_request to call layered_cycle() consistently with the POST branch, rather than calling handle_chat() directly and skipping the consciousness stack. - Export layered_cycle() from soul.elh (and dist/soul.elh) so routes.el can resolve the symbol via the header import. - Fix steward_action else branch: add explicit handling for "block" (returns safe refusal immediately, skips L3) and "redirect" (uses redirect_to field). Unknown actions now log a warning and fall back to the screened input rather than silently passing an empty string to imprint_respond(). - Document hard_bell path: clarify that omitting auto_persist/history update is intentional security isolation, and document the safety_validate second-param sentinel contract ("hard_bell" vs screen_action).Updated: layered_cycle now calls steward_session_check() which handles both behavioral fingerprinting and mission alignment in one L2 pass. Continuity status (consistent/drift/discontinuity/anomaly) is stored in session state so imprint can adjust response register. Identity anomaly and continuity concern annotations are injected into guided input before reaching L3.
Adds tests/test_layered_cycle.el — 12 integration tests covering the full L1→L2→L3→L1 stack: benign pass-through, hard-bell short-circuit, soft-bell care augmentation, steward redirect for all 5 mission-conflict signals, empty input graceful handling, sequential call isolation, and imprint state stability. Adds tests/test_layer_contract.el — contract tests verifying the JSON interface shapes between layers: safety_screen {action, content|reason|concern}, steward_align {action, content|redirect_to}, imprint_respond non-empty String, and cross-layer action propagation from L1 screen through to L1 validate.Handoff (auto) — prereq #16.
⚠ OVERLAPS #5 by ~80%: awareness.el (+402 -9), mcp-wrapper/src/main.el (+831), neuron-api.el (+173), mcp-proxy, same docs are identical to #5. This (your own consciousness-layers PR) is likely the CANONICAL version of that shared work. Decide #9 vs #5 here, merge one, rebase/close the other.
WHAT real: awareness.el +402, imprint.el +29, safety.el +33, mcp-wrapper +831, neuron-api.el +173, routes.el +40 -4, chat.el +12 -1.
REAL SOURCE: ~2,000 lines.
RISK: highest — the L0→L1→L2→L3→L1 cycle; deep cognitive-core work that genuinely needs full review, not a rubber stamp.
ORDER: the reconciliation anchor.
Pull request closed