askmy-stack · cursor · Sep 18, 2026 · Sep 18, 2026
diff --git a/.env.example b/.env.example
@@ -123,6 +123,8 @@ CORTEX_CMVK_OPENAI_MODEL=gpt-4o-mini
 CORTEX_CONTRADICTION_ENABLED=true
 CORTEX_CONTRADICTION_KAFKA=true
 CORTEX_SEMANTIC_ENABLED=false
+# Decay worker interval when Compose profile `api` is up (seconds). Default 24h.
+CORTEX_DECAY_INTERVAL_SECONDS=86400
 
 # ── OpenTelemetry (optional — enable distributed tracing) ─────────────────
 # Start collector stack: docker compose --profile observability up -d

diff --git a/ARCHITECTURE.md b/ARCHITECTURE.md
@@ -5,9 +5,9 @@
 
 ---
 
-## Current Version: v0.1 — Design Phase
-**Status:** Design only — no code written
-**Date:** 2026-05-11
+## Current Version: v1.0 — Shipped MVP
+**Status:** V1 complete (Phases 0–7). Cortex V2 (memory control plane) starts after Phase 0 audit — see `docs/CORTEX_V2.md` / `docs/CURRENT_STATE.md`.
+**Date:** 2026-09-18
 **Research foundation:** MAGMA (arXiv:2601.03236), Zep/Graphiti (arXiv:2501.13956), A-MEM (NeurIPS 2025), Field-Theoretic Memory (arXiv:2602.21220), SSGM Framework (arXiv:2603.11768)
 
 ---
@@ -792,6 +792,7 @@ Done when:
 
 | Version | Date | Changes |
 |---|---|---|
+| v1.0 | 2026-09-18 | V1 closeout — shipped MVP status; Phase 8–10 folded into V2 |
 | v0.1 | 2026-05-11 | Initial architecture design — Session 0 |
 
 ---

diff --git a/CLAUDE.md b/CLAUDE.md
@@ -93,10 +93,10 @@ LinkedIn: linkedin.com/in/abhinaysai-kamineni
 
 ## Current Phase
 
-**Phase:** 8 — Outcome tracking + coverage scoring
-**Status:** Phase 7 complete — live demo URL + README polish on `main` (set `CORTEX_API_ORIGIN` on Vercel for API-backed search)
-**Next phase:** Phase 8 deliverables (coverage scorer, outcome linker)
-**Target:** Coverage score on every query + outcome-linked decisions in the graph
+**Version:** **V1 closed** (Phases 0–7). Next work is **Cortex V2** — do not start V2 Phase 1 (data foundations) until V2 Phase 0 audit (`docs/CURRENT_STATE.md`) is merged.
+**Status:** V1 MVP shipped on `main` — connectors, pipeline, scoring, RBAC, GDPR, dashboard, MCP, live demo URL
+**Next:** V2 Phase 0 audit only (#67), then P0 Must Have (#68–#72, #77)
+**Target:** Formal V1 closeout docs + honest capability claims; V2 begins with gap report
 
 ---
 
@@ -112,9 +112,9 @@ LinkedIn: linkedin.com/in/abhinaysai-kamineni
 | 5 | Contradiction detector + decay engine | Week 3 | ✅ Done |
 | 6 | React dashboard + knowledge graph explorer | Week 3-4 | ✅ Done |
 | 7 | Live demo URL + README polish | Week 4 | ✅ Done |
-| 8 | Outcome tracking + coverage scoring | Post-launch | ⏳ |
-| 9 | Behavioral mining + elicitation bot | Post-launch | ⏳ |
-| 10 | Federated cross-org memory | v2 | ⏳ |
+| **V1** | Organizational memory MVP (phases 0–7) | — | ✅ **Closed** |
+| V2 | Memory control plane (see `docs/CORTEX_V2.md`) | — | ⏳ Phase 0 next |
+| ~~8–10~~ | Coverage / outcomes / elicitation / federation | — | Folded into V2 |
 
 ---
 

diff --git a/DECISIONS.md b/DECISIONS.md
@@ -23,6 +23,14 @@ Agent picks up OPEN instructions at session start, executes, marks DONE.
 
 ## ACTIVE INSTRUCTIONS
 
+### 2026-09-18 — Close V1 then V2 Phase 0 only
+Priority: HIGH
+Status: OPEN
+Detail:
+- V1 = phases 0–7. Do not treat coverage/outcomes/meetings as V1 deliverables.
+- After V1 closeout merges: execute V2 Phase 0 (#67) only — `docs/CORTEX_V2.md` + `docs/CURRENT_STATE.md`.
+- Do **not** start V2 Phase 1 (Claim/Evidence migrations / #68) until Phase 0 is merged.
+
 ### 2026-06-10 — LLM-backed CMVK verifiers (production)
 Priority: HIGH
 Status: DONE — `CORTEX_CMVK_BACKEND=openai|ollama` (2026-06-10)
@@ -283,6 +291,15 @@ access_policy: {
 
 ---
 
+### D-018 — 2026-09-18 — Close V1 at phases 0–7; defer remainder to V2
+**Status:** Active
+**Decision:** Cortex **V1** is complete at build phases 0–7 (connectors through live demo URL + README polish). Coverage scoring, outcome linking, meeting connectors, elicitation, and federation are **not** V1 — they move to the Cortex V2 roadmap (memory control plane). V1 closeout includes cache invalidation on graph writes (#33) and scheduled decay-worker (#36), plus README/ARCHITECTURE honesty.
+**Rationale:** README overclaimed Phase 8 features as shipped. Closing V1 on what actually works enables an honest V2 Phase 0 audit without pretending coverage/outcomes exist.
+**Alternatives rejected:** Keep calling Phase 8 "next" as V1 incomplete; ship stub `coverage_score` just to match README.
+**Owner:** Abhinaysai
+
+---
+
 ## PENDING DECISIONS (need resolution before build)
 
 | # | Decision needed | Options | Deadline | Status |
@@ -291,4 +308,4 @@ access_policy: {
 | P-002 | Local Slack message storage for testing | Real Slack workspace / Slack test fixture files | Before Phase 1 | Open |
 | P-003 | Neo4j hosting for production | Neo4j AuraDB free tier / Self-hosted EC2 | Before Phase 7 | Open |
 | P-004 | Dashboard visualization library | D3.js (full control) / React Flow (faster) | Before Phase 6 | Open |
-| P-005 | Meeting transcript connector | Recall.ai / AssemblyAI / Whisper self-hosted | Before Phase 8 | Open |
+| P-005 | Meeting transcript connector | Recall.ai / AssemblyAI / Whisper self-hosted | Before V2 | Open (see #64) |
diff --git a/README.md b/README.md
@@ -90,19 +90,19 @@ When any agent touches the payments service, Cortex enriches its context automat
 
 | Capability | Description |
 |---|---|
-| **Decision capture** | Extracts structured decisions from Slack, GitHub, Jira, Linear, meetings |
-| **Knowledge graph** | Neo4j graph: Decision → Person → System → Exception → Outcome |
+| **Decision capture** | Extracts structured decisions from Slack, GitHub, Jira, Linear |
+| **Knowledge graph** | Neo4j graph: Decision → Person → System → Exception (+ Outcome schema reserved) |
 | **Active injection** | Pushes relevant context to agents before they act — not after they ask |
 | **MCP server** | Native MCP endpoint — any Claude, Cursor, or MCP agent gets memory in one config line |
 | **Importance scoring** | Filters noise at ingestion — only signal reaches the graph |
 | **Trust scoring** | Bayesian confidence per memory node — bad inputs don't corrupt memory |
 | **Contradiction detection** | Flags when new events conflict with existing memory — no silent overwrites |
-| **Memory decay** | Old memory compresses and archives on a principled schedule |
-| **Coverage scoring** | Per-domain completeness estimate — agents know when memory is thin |
+| **Memory decay** | Batch decay engine; scheduled via `decay-worker` when the `api` Compose profile is up |
 | **RBAC** | Graph-level access control — contractors don't see salary decisions |
-| **Outcome tracking** | Links decisions to real metrics — memory becomes self-correcting |
 | **GDPR erasure** | Cascade delete with audit trail; query cache invalidated per workspace |
 
+**V1 scope note:** Coverage scoring, outcome linking, and meeting connectors are **not shipped in V1** — they are tracked for Cortex V2 (see `docs/CORTEX_V2.md` after Phase 0). Schema stubs for Outcome/coverage indices exist for forward compatibility.
+
 ---
 
 ## Production hardening (launch-ready)
@@ -111,7 +111,8 @@ These behaviors matter when Cortex runs behind auth in preview or production:
 
 | Concern | Behavior |
 |---|---|
-| **GDPR erasure** | `POST /gdpr/erase` bumps a per-workspace Redis cache epoch — stale PII cannot be served from `/query` for up to 60s |
+| **GDPR erasure** | `POST /gdpr/erase` bumps a per-workspace Redis cache epoch — stale PII cannot be served from `/query` |
+| **Pipeline writes** | Successful graph writes bump the same cache epoch — new decisions are not hidden behind a 60s stale cache |
 | **Dashboard proxy** | nginx on `:3000` forwards `/gdpr` (and `/query`, `/inject`, …) to the API — same-origin demos work |
 | **Demo smoke test** | `scripts/demo.sh` sources `.env` and sends `Authorization` when `CORTEX_DEMO_API_KEY` or `CORTEX_API_KEYS` is set |
 | **Pipeline retries** | Transient Neo4j errors do not commit Kafka offsets or land in DLQ — messages are redelivered |
@@ -144,7 +145,7 @@ After deploy, verify end-to-end: `./scripts/verify_free_deploy.sh --api https://
 ```
 ┌──────────────────────────────────────────────────────────────┐
 │  CAPTURE LAYER                                                │
-│  Slack · GitHub · Jira · Linear · Meetings · CI/CD          │
+│  Slack · GitHub · Jira · Linear  (Meetings / CI/CD → V2)    │
 │  Real-time event streams via webhooks + OAuth connectors     │
 └───────────────────────────────┬──────────────────────────────┘
                                 │ Kafka
@@ -161,19 +162,18 @@ After deploy, verify end-to-end: `./scripts/verify_free_deploy.sh --api https://
 │  Episodic    → TimescaleDB  what happened and when           │
 │  Semantic    → Qdrant       what things mean                 │
 │  Structural  → Neo4j        relationships + causal chains    │
-│  Procedural  → Neo4j        how things are done              │
 │  Hot cache   → Redis        <50ms retrieval for live agents  │
 └───────────────────────────────┬──────────────────────────────┘
                                 │
 ┌───────────────────────────────▼──────────────────────────────┐
 │  INTELLIGENCE LAYER                                           │
 │  Contradiction detector · Decay engine · Trust scorer       │
-│  Coverage scorer · Outcome linker · RBAC enforcer           │
+│  RBAC enforcer  (Coverage / Outcome linker → V2)            │
 └───────────────────────────────┬──────────────────────────────┘
                                 │
 ┌───────────────────────────────▼──────────────────────────────┐
 │  CONTEXT API                                                  │
-│  MCP server   · REST API · Python SDK · TypeScript SDK      │
+│  MCP server   · REST API · Python SDK                       │
 │  cortex.query() · cortex.inject() · cortex.remember()       │
 └──────────────────────────────────────────────────────────────┘
 ```
@@ -331,12 +331,12 @@ cortex/
 │   ├── jira/
 │   └── linear/
 ├── extraction/           # Decision extractor, entity resolver, classifier
-├── scoring/              # Importance scorer, trust scorer, coverage scorer
+├── scoring/              # Importance scorer, trust scorer, CMVK
 ├── graph/                # Neo4j schema, migrations, Cypher queries
 │   └── migrations/       # V001__initial_schema.cypher, etc.
 ├── pipeline/             # Kafka extraction worker (raw → graph)
-├── memory/               # Episodic (Timescale) + semantic (Qdrant) helpers
-├── intelligence/         # Contradiction detector, decay engine, outcome linker
+├── memory/               # Episodic (Timescale) + semantic (Qdrant) + cache epoch
+├── intelligence/         # Contradiction detector, decay engine
 ├── api/                  # FastAPI application
 ├── mcp/                  # MCP server (TypeScript)
 ├── sdk/                  # Python client (query, inject, remember)
@@ -388,9 +388,9 @@ cortex/
 | Phase 5 | Contradiction detector + decay engine | ✅ Shipped |
 | Phase 6 | React dashboard (Ask, memory map, guide, agent inject) | ✅ Shipped |
 | Phase 7 | Live demo URL + README polish | ✅ Done (wire `CORTEX_API_ORIGIN` for API-backed search) |
-| Phase 8 | Outcome tracking + coverage scoring | ⏳ Post-launch |
-| Phase 9 | Elicitation bot (implicit knowledge) | ⏳ Post-launch |
-| Phase 10 | Federated cross-org memory | ⏳ v2 |
+| **V1** | Phases 0–7 — organizational memory MVP | ✅ **Closed** |
+| V2 | Memory control plane (reliability gate, evidence, temporal, firewall) | ⏳ See `docs/CORTEX_V2.md` / issue #78 |
+| ~~Phase 8–10~~ | Coverage / outcomes / elicitation / federation | Folded into **Cortex V2** roadmap |
 
 **CI:** GitHub Actions runs `pytest` + seed dry-run on push/PR ([`.github/workflows/ci.yml`](.github/workflows/ci.yml)).
 

diff --git a/SESSIONS.md b/SESSIONS.md
@@ -601,3 +601,29 @@
 1. Set `CORTEX_API_ORIGIN` on Vercel after Render API deploy (Plan A)
 2. Implement coverage scorer (`coverage_score` on `/query`)
 3. Implement outcome linker (schema already in V004)
+
+---
+
+## Session — 2026-09-18 — V1 closeout polish
+**Duration:** ~1h
+**Phase:** V1 close (phases 0–7) before Cortex V2 Phase 0
+
+### Built
+- Merged PR #59 (demo video / launch announcement removal)
+- **#33:** `memory/cache_epoch.py` + `GraphWriter` bumps Redis cache epoch on successful writes
+- **#36:** Compose `decay-worker` under `api` profile (`CORTEX_DECAY_INTERVAL_SECONDS`)
+- README / ARCHITECTURE honesty — drop overclaimed meetings, coverage, outcomes as shipped
+- `docs/V1_RELEASE.md` — formal V1 close; Phase 8–10 folded into V2
+- CLAUDE.md: V1 closed; next is V2 Phase 0 only (no V2 Phase 1 yet)
+
+### State at end
+- V1 MVP formally closed pending this PR merge
+- Next: V2 Phase 0 audit only (#67) — CURRENT_STATE.md + CORTEX_V2.md
+
+### Decisions made
+- D-018 — Close V1 at phases 0–7; defer coverage/outcomes/meetings to V2
+
+### Next session starts with
+1. Merge V1 closeout PR
+2. V2 Phase 0 only: `docs/CORTEX_V2.md` + `docs/CURRENT_STATE.md` (#67)
+3. Do **not** start Claim/Evidence migrations until Phase 0 merges
diff --git a/api/memory.py b/api/memory.py
@@ -12,6 +12,10 @@
 
 from graph.gdpr import GdprErasureService
 from graph.query import GraphQueryService
+from memory.cache_epoch import (
+    bump_workspace_cache_epoch,
+    workspace_cache_epoch_key,
+)
 from memory.episodic import purge_raw_events_for_person
 from memory.semantic import (
     delete_decision_vectors,
@@ -22,8 +26,6 @@
 
 log = structlog.get_logger(__name__)
 
-_WORKSPACE_CACHE_EPOCH_PREFIX = "cortex:ws:"
-
 
 class MemoryService:
     """Coordinates graph reads with Redis caching and optional semantic search."""
@@ -54,20 +56,15 @@ def _build_redis_client() -> Any | None:
             return None
 
     def _workspace_cache_epoch(self, workspace_id: str) -> int:
-        """Per-workspace cache generation — bumped on GDPR erasure."""
+        """Per-workspace cache generation — bumped on GDPR erase and graph writes."""
         if self._redis is None:
             return 0
-        key = f"{_WORKSPACE_CACHE_EPOCH_PREFIX}{workspace_id}:cache_epoch"
-        raw = self._redis.get(key)
+        raw = self._redis.get(workspace_cache_epoch_key(workspace_id))
         return int(raw) if raw else 0
 
     def invalidate_workspace_cache(self, workspace_id: str) -> None:
-        """Drop cached query results for a workspace (e.g. after GDPR erasure)."""
-        if self._redis is None:
-            return
-        key = f"{_WORKSPACE_CACHE_EPOCH_PREFIX}{workspace_id}:cache_epoch"
-        self._redis.incr(key)
-        log.info("memory.cache.invalidated", workspace_id=workspace_id)
+        """Drop cached query results for a workspace (GDPR erase or after writes)."""
+        bump_workspace_cache_epoch(workspace_id, redis_client=self._redis)
 
     def _cache_key(self, prefix: str, payload: dict[str, Any]) -> str:
         workspace_id = str(payload.get("workspace_id", ""))

diff --git a/docker-compose.yml b/docker-compose.yml
@@ -341,6 +341,37 @@ services:
         condition: service_healthy
     profiles: ["api"]
 
+  decay-worker:
+    build:
+      context: .
+      dockerfile: api/Dockerfile
+    container_name: cortex-decay-worker
+    restart: unless-stopped
+    networks: [cortex]
+    env_file: .env
+    environment:
+      NEO4J_URI: bolt://neo4j:7687
+      PYTHONPATH: /app
+      # Interval between decay batches (seconds). Default 24h for local/demo.
+      CORTEX_DECAY_INTERVAL_SECONDS: ${CORTEX_DECAY_INTERVAL_SECONDS:-86400}
+    volumes:
+      - ./:/app
+    depends_on:
+      neo4j:
+        condition: service_healthy
+    # Periodic batch: run decay engine then sleep. Uses api image (Python deps).
+    entrypoint: ["/bin/sh", "-c"]
+    command:
+      - |
+        set -e
+        INTERVAL="$${CORTEX_DECAY_INTERVAL_SECONDS:-86400}"
+        echo "cortex-decay-worker: interval=$${INTERVAL}s"
+        while true; do
+          python -m intelligence.decay_engine || echo "decay_engine failed; will retry after sleep"
+          sleep "$${INTERVAL}"
+        done
+    profiles: ["api"]
+
   api:
     build:
       context: .

diff --git a/docs/V1_RELEASE.md b/docs/V1_RELEASE.md
@@ -0,0 +1,44 @@
+# Cortex V1 — Release Closeout
+
+**Status:** Closed  
+**Date:** 2026-09-18  
+**Scope:** Build phases 0–7 (organizational memory MVP)
+
+## What V1 is
+
+Cortex V1 is a working **organizational memory** system:
+
+- Capture decisions from Slack, GitHub, Jira, Linear → Kafka
+- Extract structured `DecisionEvent`s → importance / trust / CMVK write gates
+- Persist to Neo4j with graph-level RBAC; optional Qdrant semantic merge
+- Query / inject / remember via FastAPI + MCP
+- Contradiction detection, decay batch job (+ Compose `decay-worker`), GDPR erase
+- React dashboard + seeded demo + live Vercel URL
+
+## What V1 is not (deferred to V2)
+
+| Claim often seen in older docs | Reality |
+|---|---|
+| Coverage scoring on `/query` | Schema/UI stub only — V2 / #62 / #76 |
+| Outcome linking to metrics | Schema only (V004) — V2 / #63 / #75 |
+| Meeting / CI/CD connectors | Not implemented — V2 / #64 |
+| Memory Reliability Gate | V2 P0 / #71 |
+| Evidence Graph | V2 P0 / #68–#69 |
+| Procedural memory / firewall | V2 P1 / #73–#74 |
+
+## V1 closeout fixes (this release)
+
+- Redis query-cache epoch bumped on **pipeline graph writes** (not only GDPR) — #33
+- `decay-worker` scheduled service under Compose `api` profile — #36
+- README / ARCHITECTURE honesty pass (no overclaimed coverage/outcomes/meetings)
+- Formal V1 closed; Phase 8–10 folded into Cortex V2 roadmap (#78)
+
+## Ops remaining (not blockers for V1 close)
+
+- Set `CORTEX_API_ORIGIN` on Vercel after Render/Plan A API deploy (#61)
+- Public webhook URLs for live connectors (optional for demos; inject scripts work)
+
+## Next
+
+1. **V2 Phase 0 only:** `docs/CORTEX_V2.md` + `docs/CURRENT_STATE.md` (#67)  
+2. Do **not** start V2 Phase 1 (Claim/Evidence migrations) until Phase 0 merges.
diff --git a/graph/writer.py b/graph/writer.py
@@ -26,6 +26,7 @@
 from neo4j import Driver, GraphDatabase
 
 from graph.rbac import serialize_access_policy
+from memory.cache_epoch import bump_workspace_cache_epoch
 from scoring.trust_scorer import is_writable
 from scoring.write_pipeline import assert_scored_for_write
 from shared.models import IMPORTANCE_DISCARD, DecisionEvent
@@ -227,6 +228,9 @@ def write(
                 valid_at=valid_at,
             )
 
+        # Invalidate Redis query/inject cache so new decisions are visible immediately.
+        bump_workspace_cache_epoch(decision.workspace_id)
+
         log.info(
             "graph.write.success",
             event_id=decision.event_id,