---
id: 9ea72015-0b85-4a42-9ae1-144866f8d86f
type: solution
title: "Cognitive Memory v3.0: Rich Edges + Hybrid Embeddings + MCP Server"
tags: [cognitive-memory, mcp, architecture, upgrade]
importance: 0.9
confidence: 0.8
created: "2026-02-19T20:10:42.128691+00:00"
updated: "2026-03-03T05:32:47.939579+00:00"
relations:
  - target: dc6b6c6f-1319-47cf-a691-dd40751ff121
    type: BUILDS_ON
    direction: incoming
    strength: 0.85
    edge_id: 78728fca-82c7-4c30-b677-aadc663fb40b
  - target: 89f32ef3-2d87-4b8c-aa6e-ce5a7f60566c
    type: BUILDS_ON
    direction: incoming
    strength: 0.8
    edge_id: 6100de8d-c1dc-4ec9-8981-5c65a396da0e
  - target: b1ba1b27-d2bf-44f9-a1f5-0ace2017dc18
    type: BUILDS_ON
    direction: incoming
    strength: 0.8
    edge_id: f5227880-ea2c-4330-b7f1-eb367fc12c9e
  - target: dc9c2638-360e-4e15-990c-76fcd16672c0
    type: FOLLOWS
    direction: incoming
    strength: 0.8
    edge_id: 30dd2fb7-f75f-4083-8b7c-52b8ce95cfc8
  - target: a41644f1-2dd1-4706-8fc8-e2242c1702cb
    type: FOLLOWS
    direction: incoming
    strength: 0.85
    edge_id: aea53b08-79a6-446a-bd1c-13f37b872c4e
  - target: 089f4cd8-4b2b-49a5-aac1-af2811b8ea7d
    type: RELATED_TO
    direction: incoming
    strength: 0.8
    edge_id: 2f7fc7c4-66b6-44d1-be62-2ba56af9b687
  - target: 8023fae6-eec1-4c93-bac8-1bd85dd61bae
    type: BUILDS_ON
    direction: incoming
    strength: 0.76
    edge_id: 454fe3a6-e243-4cc0-bb9a-f12426f8deb3
  - target: ee85bf89-93f0-48e4-b911-7ddc18a735b5
    type: BUILDS_ON
    direction: incoming
    strength: 0.85
    edge_id: bc703713-c85f-478c-be2c-0f0cd35f53c7
  - target: ab3ffc18-1707-42ff-ad7e-bffb16f6fdef
    type: BUILDS_ON
    direction: incoming
    strength: 0.75
    edge_id: 13e98a58-b81d-49b0-89b9-0f23d3d24657
  - target: 21be13ae-c01b-4bf1-917a-9e1ef18312a8
    type: RELATED_TO
    direction: incoming
    strength: 0.6
    edge_id: 776d09a5-d1b3-4fe3-ac8a-5b41e40460d1
  - target: 2d6f02bc-5a7c-404d-a566-36107aaaf6e6
    type: RELATED_TO
    direction: incoming
    strength: 0.82
    edge_id: f30f5536-be57-4afc-ad64-bb4e2b84bc19
  - target: e301095c-7d4c-45d8-8352-9491e4ad6603
    type: BUILDS_ON
    direction: incoming
    strength: 0.73
    edge_id: 2eee0082-215a-41b9-90bc-ea2d757138ff
  - target: 96a6c125-bb69-45af-b1b6-a9b48654fb7e
    type: BUILDS_ON
    direction: incoming
    strength: 0.82
    edge_id: 255bd837-327a-4889-a60b-ea5afaa32a11
  - target: 1af571c5-cf7a-4e40-b324-6973b56f2502
    type: RELATED_TO
    direction: incoming
    strength: 0.81
    edge_id: 33a0bb8c-c614-48cb-ae31-d7cab47d811c
---

Major upgrade to cognitive-memory skill. Phase 1: Rich edges as first-class markdown files in graph/edges/ with bidirectional frontmatter refs (edge_id field). relate() returns edge_id string instead of bool. Cascade deletion on memory delete. CLI: edge-get, edge-search, edge-update, edge-delete. Phase 2: Hybrid embedding providers (Ollama local + OpenAI optional) with automatic fallback chain. _config.json stores provider settings (gitignored). Dimension mismatch safety triggers re-embedding on provider switch. Phase 3: MCP server (mcp_server.py) with 18 tools via JSON-RPC 2.0 stdio protocol. Registered in ~/.claude.json. Phase 4: Updated SKILL.md, SCHEMA.md, feature.json to v3.0.0. Index version bumped from 1 to 2 (adds edges section). All stdlib-only, no external dependencies.