release: v3.12.0 — upstream MCP ingest (ingest_codebase + ingest_prd)

Highlights since v3.11.0:

- **New: `ingest_codebase`** — pulls tree-sitter + graph analysis from
  the ai-automatised-pipeline MCP server into Cortex memories +
  knowledge-graph entities/edges + wiki reference pages. Graph-path
  memoised per project for idempotent re-runs.
- **New: `ingest_prd`** — pulls a PRD from disk / inline markdown /
  prd-gen pipeline state into a wiki spec page + per-bullet
  decision/requirement memories. Optional `validate=true` routes
  through prd-gen's `validate_prd_document` for quality signals.
- **Deleted: `run_pipeline`** — ai-architect is dead. Cortex was
  driving a non-existent server; replaced with the consumer-side
  pattern above.
- ADR-0044 documents the upstream-consumer design.
- `docs/mcp-connections.example.json` ships as a user template.

Tool count: 41 → 42.
Test count: 2087 → 2098 (+11 ingest tests). All pass.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: cdeust

.claude-plugin/marketplace.json

@@ -6,14 +6,14 @@
   },
   "metadata": {
     "description": "Persistent memory and cognitive profiling plugins for Claude Code",
-    "version": "3.11.0"
+    "version": "3.12.0"
   },
   "plugins": [
     {
       "name": "cortex",
       "source": "./",
-      "description": "Persistent memory and cognitive profiling for Claude Code — thermodynamic memory with heat/decay, intent-aware retrieval, biological plasticity, codebase intelligence, and cognitive profiling. 41 MCP tools with enriched schemas. PostgreSQL + pgvector in CLI mode; automatic SQLite fallback in Cowork/sandboxed mode. Curated wiki (ADRs, specs, lessons) with audit-artefact filtering. Consolidate is now set-based SQL batched — decay/plasticity/pruning run 100-500× faster on large stores. Docker image available.",
-      "version": "3.11.0",
+      "description": "Persistent memory and cognitive profiling for Claude Code — thermodynamic memory with heat/decay, intent-aware retrieval, biological plasticity, codebase intelligence, and cognitive profiling. 42 MCP tools with enriched schemas. PostgreSQL + pgvector in CLI mode; automatic SQLite fallback in Cowork/sandboxed mode. Curated wiki (ADRs, specs, lessons) with audit-artefact filtering. Consolidate is set-based SQL batched — decay/plasticity/pruning run 100-500× faster on large stores. Ingests codebase analysis (ai-automatised-pipeline) and PRDs (prd-spec-generator) into wiki + memory + knowledge graph. Docker image available.",
+      "version": "3.12.0",
       "author": {
         "name": "Clement Deust",
         "email": "admin@ai-architect.tools"

.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "cortex",
   "description": "Persistent memory for Claude Code — remembers across sessions automatically. Install and forget. Scientific retrieval backed by 41 published papers.",
-  "version": "3.11.0",
+  "version": "3.12.0",
   "author": {
     "name": "Clement Deust",
     "email": "admin@ai-architect.tools"

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "neuro-cortex-memory"
-version = "3.11.0"
+version = "3.12.0"
 description = "Scientifically-grounded memory system based on computational neuroscience research"
 readme = "README.md"
 license = "MIT"