audit: clean Docker architecture, consolidate test fixtures, add containerized test runner (#94)

Dockerfile.ollama

@@ -1,49 +0,0 @@
-# ── Ollama with Pre-loaded Models ──────────────────────────────────────────────
-#
-# This Dockerfile extends the official Ollama image with pre-loaded models
-# for faster startup and better performance.
-#
-# Build: docker build -f Dockerfile.ollama -t timmy-ollama:latest .
-# Run:   docker run -p 11434:11434 -v ollama-data:/root/.ollama timmy-ollama:latest
-
-FROM ollama/ollama:latest
-
-# Set environment variables
-ENV OLLAMA_HOST=0.0.0.0:11434
-
-# Create a startup script that pulls models on first run
-RUN mkdir -p /app
-COPY <<EOF /app/init-models.sh
-#!/bin/bash
-set -e
-
-echo "🚀 Ollama startup — checking for models..."
-
-# Start Ollama in the background
-ollama serve &
-OLLAMA_PID=$!
-
-# Wait for Ollama to be ready
-echo "⏳ Waiting for Ollama to be ready..."
-for i in {1..30}; do
-  if curl -s http://localhost:11434/api/tags > /dev/null 2>&1; then
-    echo "✓ Ollama is ready"
-    break
-  fi
-  sleep 1
-done
-
-# Pull the default model if not already present
-echo "📥 Pulling llama3.2 model..."
-ollama pull llama3.2 || true
-
-echo "✓ Ollama initialization complete"
-
-# Keep the process running
-wait $OLLAMA_PID
-EOF
-
-RUN chmod +x /app/init-models.sh
-
-# Use the init script as the entrypoint
-ENTRYPOINT ["/app/init-models.sh"]

Makefile

@@ -1,7 +1,9 @@
 .PHONY: install install-bigbrain dev nuke fresh test test-cov test-cov-html watch lint clean help \
         up down logs \
         docker-build docker-up docker-down docker-agent docker-logs docker-shell \
-        cloud-deploy cloud-up cloud-down cloud-logs cloud-status cloud-update
+        test-docker test-docker-cov test-docker-functional test-docker-build test-docker-down \
+        cloud-deploy cloud-up cloud-down cloud-logs cloud-status cloud-update \
+        logs-up logs-down logs-kibana
 
 PYTEST      := poetry run pytest
 UVICORN     := poetry run uvicorn
@@ -114,6 +116,38 @@ test-cov-html:
 test-ollama:
 	FUNCTIONAL_DOCKER=1 $(PYTEST) tests/functional/test_ollama_chat.py -v --tb=long -x
 
+# ── Docker test containers ───────────────────────────────────────────────────
+# Clean containers from cached images; source bind-mounted for fast iteration.
+# Rebuild only needed when pyproject.toml / poetry.lock change.
+
+# Build the test image (cached — fast unless deps change)
+test-docker-build:
+	DOCKER_BUILDKIT=1 docker compose -f docker-compose.test.yml build
+
+# Run all unit + integration tests in a clean container (default)
+# Override: make test-docker ARGS="-k swarm -v"
+test-docker: test-docker-build
+	docker compose -f docker-compose.test.yml run --rm test \
+	    pytest tests/ -q --tb=short $(ARGS)
+	docker compose -f docker-compose.test.yml down -v
+
+# Run tests with coverage inside a container
+test-docker-cov: test-docker-build
+	docker compose -f docker-compose.test.yml run --rm test \
+	    pytest tests/ --cov=src --cov-report=term-missing -q $(ARGS)
+	docker compose -f docker-compose.test.yml down -v
+
+# Spin up the full stack (dashboard + optional Ollama) and run functional tests
+test-docker-functional: test-docker-build
+	docker compose -f docker-compose.test.yml --profile functional up -d --wait
+	docker compose -f docker-compose.test.yml run --rm test \
+	    pytest tests/functional/ -v --tb=short $(ARGS) || true
+	docker compose -f docker-compose.test.yml --profile functional down -v
+
+# Tear down any leftover test containers and volumes
+test-docker-down:
+	docker compose -f docker-compose.test.yml --profile functional --profile ollama --profile agents down -v
+
 # ── Code quality ──────────────────────────────────────────────────────────────
 
 lint:
@@ -226,6 +260,22 @@ cloud-scale:
 cloud-pull-model:
 	docker exec timmy-ollama ollama pull $${MODEL:-llama3.2}
 
+# ── ELK Logging ──────────────────────────────────────────────────────────────
+# Overlay on top of the production stack for centralised log aggregation.
+# Kibana UI: http://localhost:5601
+
+logs-up:
+	docker compose -f docker-compose.prod.yml -f docker-compose.logging.yml up -d
+
+logs-down:
+	docker compose -f docker-compose.prod.yml -f docker-compose.logging.yml down
+
+logs-kibana:
+	@echo "Opening Kibana at http://localhost:5601 ..."
+	@command -v open >/dev/null 2>&1 && open http://localhost:5601 || \
+	 command -v xdg-open >/dev/null 2>&1 && xdg-open http://localhost:5601 || \
+	 echo "  → Open http://localhost:5601 in your browser"
+
 # ── Housekeeping ──────────────────────────────────────────────────────────────
 
 clean:
@@ -268,6 +318,15 @@ help:
 	@echo "  make pre-commit-install install pre-commit hooks"
 	@echo "  make clean            remove build artefacts and caches"
 	@echo ""
+	@echo "  Docker Testing (Clean Containers)"
+	@echo "  ─────────────────────────────────────────────────"
+	@echo "  make test-docker              run tests in clean container"
+	@echo "  make test-docker ARGS=\"-k swarm\" filter tests in container"
+	@echo "  make test-docker-cov          tests + coverage in container"
+	@echo "  make test-docker-functional   full-stack functional tests"
+	@echo "  make test-docker-build        build test image (cached)"
+	@echo "  make test-docker-down         tear down test containers"
+	@echo ""
 	@echo "  Docker (Advanced)"
 	@echo "  ─────────────────────────────────────────────────"
 	@echo "  make docker-build     build the timmy-time:latest image"
@@ -289,3 +348,9 @@ help:
 	@echo "  make cloud-scale N=4  scale agent workers"
 	@echo "  make cloud-pull-model MODEL=llama3.2  pull LLM model"
 	@echo ""
+	@echo "  ELK Log Aggregation"
+	@echo "  ─────────────────────────────────────────────────"
+	@echo "  make logs-up          start prod + ELK stack"
+	@echo "  make logs-down        stop prod + ELK stack"
+	@echo "  make logs-kibana      open Kibana UI (http://localhost:5601)"
+	@echo ""

deploy/elk/elasticsearch.yml

@@ -0,0 +1,23 @@
+# ── Elasticsearch — Single-node config for Timmy Time ───────────────────────
+#
+# Minimal config for a single-node deployment.  For multi-node clusters
+# see: https://www.elastic.co/guide/en/elasticsearch/reference/current/docker.html
+
+cluster.name: timmy-logs
+node.name: timmy-es-01
+
+# Single-node discovery (no cluster formation overhead)
+discovery.type: single-node
+
+# Bind to all interfaces inside the container
+network.host: 0.0.0.0
+
+# Security: disable X-Pack security for internal-only deployments.
+# Enable and configure TLS if exposing Elasticsearch externally.
+xpack.security.enabled: false
+
+# Memory: let the JVM use container-aware defaults
+# (set ES_JAVA_OPTS in docker-compose for explicit heap)
+
+# Index lifecycle — auto-delete old logs after 30 days
+# Applied via ILM policy created by Logstash on first boot.

deploy/elk/kibana.yml

@@ -0,0 +1,14 @@
+# ── Kibana — Dashboard config for Timmy Time ────────────────────────────────
+
+server.name: timmy-kibana
+server.host: "0.0.0.0"
+server.port: 5601
+
+# Connect to Elasticsearch on the Docker network
+elasticsearch.hosts: ["http://elasticsearch:9200"]
+
+# Disable telemetry (sovereign deployment)
+telemetry.enabled: false
+
+# Default index pattern
+# Kibana will auto-create this on first boot when logs arrive.

deploy/elk/logstash.conf

@@ -0,0 +1,78 @@
+# ── Logstash Pipeline — Timmy Time log aggregation ──────────────────────────
+#
+# Collects Docker container logs via the GELF input, parses them,
+# and ships structured events to Elasticsearch.
+#
+# Flow:  Docker (GELF driver) → Logstash :12201/udp → Elasticsearch
+
+input {
+  # GELF (Graylog Extended Log Format) — Docker's native structured log driver.
+  # Each container sends logs here automatically via the logging driver config
+  # in docker-compose.logging.yml.
+  gelf {
+    port => 12201
+    type => "docker"
+  }
+}
+
+filter {
+  # ── Tag by container name ──────────────────────────────────────────────────
+  # Docker GELF driver populates these fields automatically:
+  #   container_name, container_id, image_name, tag, command, created
+
+  # Strip leading "/" from container_name (Docker convention)
+  if [container_name] {
+    mutate {
+      gsub => ["container_name", "^/", ""]
+    }
+  }
+
+  # ── Parse JSON log lines (FastAPI/uvicorn emit JSON when configured) ──────
+  if [message] =~ /^\{/ {
+    json {
+      source => "message"
+      target => "log"
+      skip_on_invalid_json => true
+    }
+  }
+
+  # ── Extract log level ─────────────────────────────────────────────────────
+  # Try structured field first, fall back to regex on raw message
+  if [log][level] {
+    mutate { add_field => { "log_level" => "%{[log][level]}" } }
+  } else if [level] {
+    mutate { add_field => { "log_level" => "%{level}" } }
+  } else {
+    grok {
+      match => { "message" => "(?i)(?<log_level>DEBUG|INFO|WARNING|ERROR|CRITICAL)" }
+      tag_on_failure => []
+    }
+  }
+
+  # Normalise to uppercase
+  if [log_level] {
+    mutate { uppercase => ["log_level"] }
+  }
+
+  # ── Add service metadata ──────────────────────────────────────────────────
+  mutate {
+    add_field => { "environment" => "production" }
+    add_field => { "project"     => "timmy-time" }
+  }
+}
+
+output {
+  elasticsearch {
+    hosts => ["http://elasticsearch:9200"]
+    index => "timmy-logs-%{+YYYY.MM.dd}"
+
+    # ILM policy: auto-rollover + delete after 30 days
+    ilm_enabled     => true
+    ilm_rollover_alias => "timmy-logs"
+    ilm_pattern     => "{now/d}-000001"
+    ilm_policy      => "timmy-logs-policy"
+  }
+
+  # Also print to stdout for debugging (disable in production)
+  # stdout { codec => rubydebug }
+}

docker-compose.enhanced.yml

@@ -1,130 +0,0 @@
-# ── Timmy Time — Enhanced Docker Compose with Ollama ──────────────────────────
-#
-# This enhanced version includes Ollama service for local LLM inference.
-# Services:
-#   ollama      Local LLM inference server (required for Timmy)
-#   dashboard   FastAPI app + swarm coordinator
-#   timmy       Timmy sovereign agent
-#   agent       Swarm worker template (scale with --scale agent=N --profile agents)
-#
-# Usage:
-#   docker compose -f docker-compose.enhanced.yml up -d
-#   docker compose -f docker-compose.enhanced.yml logs -f dashboard
-#   docker compose -f docker-compose.enhanced.yml down
-
-services:
-
-  # ── Ollama — Local LLM Inference Server ────────────────────────────────────
-  ollama:
-    image: ollama/ollama:latest
-    container_name: timmy-ollama
-    ports:
-      - "11434:11434"
-    volumes:
-      - ollama-data:/root/.ollama
-    environment:
-      OLLAMA_HOST: "0.0.0.0:11434"
-    networks:
-      - swarm-net
-    restart: unless-stopped
-    healthcheck:
-      test: ["CMD", "curl", "-f", "http://localhost:11434/api/tags"]
-      interval: 30s
-      timeout: 5s
-      retries: 3
-      start_period: 30s
-
-  # ── Dashboard (coordinator + FastAPI) ──────────────────────────────────────
-  dashboard:
-    build: .
-    image: timmy-time:latest
-    container_name: timmy-dashboard
-    user: "0:0"
-    ports:
-      - "8000:8000"
-    volumes:
-      - timmy-data:/app/data
-      - ./src:/app/src
-      - ./static:/app/static
-    environment:
-      DEBUG: "true"
-      # Point to Ollama container
-      OLLAMA_URL: "http://ollama:11434"
-      GROK_ENABLED: "${GROK_ENABLED:-false}"
-      XAI_API_KEY: "${XAI_API_KEY:-}"
-      GROK_DEFAULT_MODEL: "${GROK_DEFAULT_MODEL:-grok-3-fast}"
-    networks:
-      - swarm-net
-    depends_on:
-      ollama:
-        condition: service_healthy
-    restart: unless-stopped
-    healthcheck:
-      test: ["CMD", "curl", "-f", "http://localhost:8000/health"]
-      interval: 30s
-      timeout: 5s
-      retries: 3
-      start_period: 15s
-
-  # ── Timmy — Sovereign AI Agent ─────────────────────────────────────────────
-  timmy:
-    build: .
-    image: timmy-time:latest
-    container_name: timmy-agent
-    volumes:
-      - timmy-data:/app/data
-      - ./src:/app/src
-    environment:
-      COORDINATOR_URL: "http://dashboard:8000"
-      OLLAMA_URL: "http://ollama:11434"
-      TIMMY_AGENT_ID: "timmy"
-    command: ["python", "-m", "timmy.docker_agent"]
-    networks:
-      - swarm-net
-    depends_on:
-      dashboard:
-        condition: service_healthy
-      ollama:
-        condition: service_healthy
-    restart: unless-stopped
-
-  # ── Agent Worker Template ──────────────────────────────────────────────────
-  # Scale: docker compose -f docker-compose.enhanced.yml up --scale agent=4 --profile agents
-  agent:
-    build: .
-    image: timmy-time:latest
-    profiles:
-      - agents
-    volumes:
-      - timmy-data:/app/data
-      - ./src:/app/src
-    environment:
-      COORDINATOR_URL: "http://dashboard:8000"
-      OLLAMA_URL: "http://ollama:11434"
-      AGENT_NAME: "${AGENT_NAME:-Worker}"
-      AGENT_CAPABILITIES: "${AGENT_CAPABILITIES:-general}"
-    command: ["sh", "-c", "python -m swarm.agent_runner --agent-id agent-$(hostname) --name $${AGENT_NAME:-Worker}"]
-    networks:
-      - swarm-net
-    depends_on:
-      dashboard:
-        condition: service_healthy
-      ollama:
-        condition: service_healthy
-    restart: unless-stopped
-
-# ── Volumes ────────────────────────────────────────────────────────────────────
-volumes:
-  timmy-data:
-    driver: local
-    driver_opts:
-      type: none
-      o: bind
-      device: "${PWD}/data"
-  ollama-data:
-    driver: local
-
-# ── Internal Network ───────────────────────────────────────────────────────────
-networks:
-  swarm-net:
-    driver: bridge

docker-compose.logging.yml

@@ -0,0 +1,127 @@
+# ── Timmy Time — ELK Log Aggregation Overlay ────────────────────────────────
+#
+# Adds Elasticsearch + Logstash + Kibana alongside the production stack.
+# Use as an overlay on top of the prod compose:
+#
+#   docker compose \
+#     -f docker-compose.prod.yml \
+#     -f docker-compose.logging.yml \
+#     up -d
+#
+# ── How it works ────────────────────────────────────────────────────────────
+#
+#   1. Every container's Docker logging driver is set to GELF, which sends
+#      structured log events (JSON with container metadata) over UDP.
+#
+#   2. Logstash listens on :12201/udp, parses the GELF messages, extracts
+#      log levels, parses JSON payloads from FastAPI/uvicorn, and adds
+#      project metadata.
+#
+#   3. Logstash ships the enriched events to Elasticsearch, indexed by day
+#      (timmy-logs-YYYY.MM.dd) with a 30-day ILM retention policy.
+#
+#   4. Kibana provides the web UI on :5601 for searching, filtering,
+#      and building dashboards over the indexed logs.
+#
+# ── Access ──────────────────────────────────────────────────────────────────
+#   Kibana:         http://localhost:5601
+#   Elasticsearch:  http://localhost:9200  (API only, not exposed by default)
+#
+# ── Resource notes ──────────────────────────────────────────────────────────
+#   Elasticsearch: ~512 MB heap (ES_JAVA_OPTS below).  Increase for
+#                  high-throughput deployments.
+#   Logstash:      ~256 MB heap.  Lightweight for GELF → ES pipeline.
+#   Kibana:        ~300 MB RAM.  Stateless — safe to restart anytime.
+#
+#   Total overhead: ~1.1 GB RAM on top of the base production stack.
+
+services:
+
+  # ── Elasticsearch — log storage and search engine ─────────────────────────
+  elasticsearch:
+    image: docker.elastic.co/elasticsearch/elasticsearch:8.17.0
+    container_name: timmy-elasticsearch
+    volumes:
+      - es-data:/usr/share/elasticsearch/data
+      - ./deploy/elk/elasticsearch.yml:/usr/share/elasticsearch/config/elasticsearch.yml:ro
+    environment:
+      ES_JAVA_OPTS: "-Xms512m -Xmx512m"
+    networks:
+      - swarm-net
+    restart: unless-stopped
+    healthcheck:
+      test: ["CMD-SHELL", "curl -sf http://localhost:9200/_cluster/health | grep -q '\"status\":\"green\\|yellow\"'"]
+      interval: 30s
+      timeout: 10s
+      retries: 5
+      start_period: 60s
+
+  # ── Logstash — log pipeline (GELF in → Elasticsearch out) ────────────────
+  logstash:
+    image: docker.elastic.co/logstash/logstash:8.17.0
+    container_name: timmy-logstash
+    volumes:
+      - ./deploy/elk/logstash.conf:/usr/share/logstash/pipeline/logstash.conf:ro
+    environment:
+      LS_JAVA_OPTS: "-Xms256m -Xmx256m"
+    ports:
+      - "12201:12201/udp"   # GELF input from Docker logging driver
+    networks:
+      - swarm-net
+    depends_on:
+      elasticsearch:
+        condition: service_healthy
+    restart: unless-stopped
+
+  # ── Kibana — log visualisation UI ─────────────────────────────────────────
+  kibana:
+    image: docker.elastic.co/kibana/kibana:8.17.0
+    container_name: timmy-kibana
+    volumes:
+      - ./deploy/elk/kibana.yml:/usr/share/kibana/config/kibana.yml:ro
+    ports:
+      - "5601:5601"
+    networks:
+      - swarm-net
+    depends_on:
+      elasticsearch:
+        condition: service_healthy
+    restart: unless-stopped
+
+  # ── Override existing services to use GELF logging driver ─────────────────
+  # These extend the services defined in docker-compose.prod.yml.
+  # Docker merges the logging config into the existing service definition.
+
+  dashboard:
+    logging:
+      driver: gelf
+      options:
+        gelf-address: "udp://localhost:12201"
+        tag: "dashboard"
+    depends_on:
+      logstash:
+        condition: service_started
+
+  timmy:
+    logging:
+      driver: gelf
+      options:
+        gelf-address: "udp://localhost:12201"
+        tag: "timmy-agent"
+    depends_on:
+      logstash:
+        condition: service_started
+
+  ollama:
+    logging:
+      driver: gelf
+      options:
+        gelf-address: "udp://localhost:12201"
+        tag: "ollama"
+    depends_on:
+      logstash:
+        condition: service_started
+
+# ── Persistent volume for Elasticsearch indices ────────────────────────────
+volumes:
+  es-data:

docker-compose.microservices.yml

@@ -11,8 +11,6 @@
 #   docker compose -f docker-compose.microservices.yml logs -f dashboard
 #   docker compose -f docker-compose.microservices.yml up --scale worker=4
 
-version: "3.9"
-
 services:
 
   # ── Ollama LLM Service ────────────────────────────────────────────────────

docker-compose.test.yml

@@ -1,54 +1,73 @@
-# ── Timmy Time — test stack ──────────────────────────────────────────────────
+# ── Timmy Time — Test Stack ──────────────────────────────────────────────────
 #
-# Lightweight compose for functional tests.  Runs the dashboard on port 18000
+# Clean containers for test runs.  Designed for fast iteration:
-# and optional agent workers on the swarm-test-net network.
+#   • Cached builder layers  — only rebuilds when pyproject.toml changes
+#   • Bind-mounted source    — code changes are instant, no rebuild needed
+#   • Ephemeral test-data    — every run starts with clean state
 #
-# Profiles:
+# ── Profiles ────────────────────────────────────────────────────────────────
-#   (default)   dashboard only (Ollama on host via host.docker.internal)
+#   (default)   test runner only (unit + integration tests)
-#   ollama      adds a containerised Ollama instance + auto model pull
+#   functional  adds a live dashboard on port 18000 for HTTP-level tests
-#   agents      adds scalable agent workers
+#   ollama      adds containerised Ollama (CPU, qwen2.5:0.5b) for LLM tests
+#   agents      adds swarm agent workers for multi-agent tests
 #
-# Usage:
+# ── Quick-start ─────────────────────────────────────────────────────────────
-#   # Swarm tests (no LLM needed):
+#   make test-docker                         # unit + integration in container
-#   FUNCTIONAL_DOCKER=1 pytest tests/functional/test_docker_swarm.py -v
+#   make test-docker ARGS="-k swarm"         # filter tests
+#   make test-docker-functional              # full-stack functional tests
+#   make test-docker-cov                     # with coverage report
 #
-#   # Full-stack with Ollama (pulls qwen2.5:0.5b automatically):
+# ── Manual usage ────────────────────────────────────────────────────────────
-#   FUNCTIONAL_DOCKER=1 pytest tests/functional/test_ollama_chat.py -v
+#   docker compose -f docker-compose.test.yml run --rm test
-#
+#   docker compose -f docker-compose.test.yml run --rm test pytest tests/swarm -v
-# Or manually:
+#   docker compose -f docker-compose.test.yml --profile functional up -d --wait
-#   docker compose -f docker-compose.test.yml -p timmy-test up -d --build --wait
+#   docker compose -f docker-compose.test.yml down -v
-#   curl http://localhost:18000/health
-#   docker compose -f docker-compose.test.yml -p timmy-test down -v
 
 services:
 
-  # ── Ollama — local LLM for functional tests ───────────────────────────────
+  # ── Test Runner ───────────────────────────────────────────────────────────
-  # Activated with: --profile ollama
+  # Runs pytest in a clean container.  Exits when tests complete.
-  # Uses a tiny model (qwen2.5:0.5b, ~400 MB) so it runs on CPU-only CI.
+  # Source and tests are bind-mounted so code changes don't require a rebuild.
-  ollama:
+  test:
-    image: ollama/ollama:latest
+    build:
-    container_name: timmy-test-ollama
+      context: .
-    profiles:
+      dockerfile: docker/Dockerfile.test
-      - ollama
+      cache_from:
+        - timmy-test:latest
+    image: timmy-test:latest
+    volumes:
+      - ./src:/app/src:ro
+      - ./tests:/app/tests:ro
+      - ./static:/app/static:ro
+      - ./pyproject.toml:/app/pyproject.toml:ro
+      - test-data:/app/data
+    environment:
+      TIMMY_TEST_MODE: "1"
+      LIGHTNING_BACKEND: "mock"
+      PYTHONDONTWRITEBYTECODE: "1"
     networks:
-      - swarm-test-net
+      - test-net
-    healthcheck:
+    # Default command — override with: docker compose run --rm test pytest <args>
-      test: ["CMD", "curl", "-f", "http://localhost:11434/api/tags"]
+    command: ["pytest", "tests/", "-q", "--tb=short"]
-      interval: 5s
-      timeout: 5s
-      retries: 20
-      start_period: 10s
 
+  # ── Dashboard — live server for functional tests ──────────────────────────
+  # Activated with: --profile functional
   dashboard:
-    build: .
+    build:
-    image: timmy-time:test
+      context: .
+      dockerfile: docker/Dockerfile.test
+      cache_from:
+        - timmy-test:latest
+    image: timmy-test:latest
+    profiles:
+      - functional
     container_name: timmy-test-dashboard
     ports:
       - "18000:8000"
     volumes:
+      - ./src:/app/src:ro
+      - ./static:/app/static:ro
       - test-data:/app/data
-      - ./src:/app/src
-      - ./static:/app/static
     environment:
       DEBUG: "true"
       TIMMY_TEST_MODE: "1"
@@ -58,7 +77,8 @@ services:
     extra_hosts:
       - "host.docker.internal:host-gateway"
     networks:
-      - swarm-test-net
+      - test-net
+    command: ["uvicorn", "dashboard.app:app", "--host", "0.0.0.0", "--port", "8000"]
     healthcheck:
       test: ["CMD", "curl", "-f", "http://localhost:8000/health"]
       interval: 5s
@@ -66,14 +86,38 @@ services:
       retries: 10
       start_period: 10s
 
+  # ── Ollama — local LLM for functional tests ──────────────────────────────
+  # Activated with: --profile ollama
+  # Uses a tiny model (qwen2.5:0.5b, ~400 MB) so it runs on CPU-only CI.
+  ollama:
+    image: ollama/ollama:latest
+    container_name: timmy-test-ollama
+    profiles:
+      - ollama
+    networks:
+      - test-net
+    healthcheck:
+      test: ["CMD", "curl", "-f", "http://localhost:11434/api/tags"]
+      interval: 5s
+      timeout: 5s
+      retries: 20
+      start_period: 10s
+
+  # ── Agent — swarm worker for multi-agent tests ───────────────────────────
+  # Activated with: --profile agents
+  # Scale: docker compose -f docker-compose.test.yml --profile agents up --scale agent=4
   agent:
-    build: .
+    build:
-    image: timmy-time:test
+      context: .
+      dockerfile: docker/Dockerfile.test
+      cache_from:
+        - timmy-test:latest
+    image: timmy-test:latest
     profiles:
       - agents
     volumes:
+      - ./src:/app/src:ro
       - test-data:/app/data
-      - ./src:/app/src
     environment:
       COORDINATOR_URL: "http://dashboard:8000"
       OLLAMA_URL: "${OLLAMA_URL:-http://host.docker.internal:11434}"
@@ -83,16 +127,21 @@ services:
       TIMMY_TEST_MODE: "1"
     extra_hosts:
       - "host.docker.internal:host-gateway"
-    command: ["sh", "-c", "python -m swarm.agent_runner --agent-id agent-$(hostname) --name $${AGENT_NAME:-TestWorker}"]
+    command: >-
+      sh -c "python -m swarm.agent_runner
+        --agent-id agent-$$(hostname)
+        --name $${AGENT_NAME:-TestWorker}"
     networks:
-      - swarm-test-net
+      - test-net
     depends_on:
       dashboard:
         condition: service_healthy
 
+# ── Ephemeral volume — destroyed with `docker compose down -v` ─────────────
 volumes:
   test-data:
 
+# ── Isolated test network ─────────────────────────────────────────────────
 networks:
-  swarm-test-net:
+  test-net:
     driver: bridge

docker-compose.yml

@@ -1,7 +1,8 @@
-# ── Timmy Time — docker-compose ─────────────────────────────────────────────
+# ── Timmy Time — Development Compose ────────────────────────────────────────
 #
 # Services
 #   dashboard   FastAPI app + swarm coordinator (always on)
+#   timmy       Sovereign AI agent (separate container)
 #   agent       Swarm worker template — scale with:
 #                 docker compose up --scale agent=N --profile agents
 #
@@ -14,6 +15,27 @@
 #   make docker-agent    add one agent worker
 #   make docker-down     stop everything
 #   make docker-logs     tail logs
+#
+# ── Security note: root user in dev ─────────────────────────────────────────
+# This dev compose runs containers as root (user: "0:0") so that
+# bind-mounted host files (./src, ./static) are readable regardless of
+# host UID/GID — the #1 cause of 403 errors on macOS.
+#
+# Production (docker-compose.prod.yml) uses NO bind mounts and runs as
+# the Dockerfile's non-root "timmy" user.  Never expose this dev compose
+# to untrusted networks.
+#
+# ── Ollama host access ──────────────────────────────────────────────────────
+# By default OLLAMA_URL points to http://host.docker.internal:11434 which
+# reaches Ollama running on the Docker host (macOS/Windows native).
+#
+# Linux:  The extra_hosts entry maps host.docker.internal → host-gateway,
+#         which resolves to the host IP on Docker 20.10+.  If you run an
+#         older Docker version, set OLLAMA_URL=http://172.17.0.1:11434
+#         in your .env file instead.
+#
+# Containerised Ollama:  Use docker-compose.microservices.yml which runs
+#         Ollama as a sibling container on the same network.
 
 services:
 
@@ -22,12 +44,7 @@ services:
     build: .
     image: timmy-time:latest
     container_name: timmy-dashboard
-    # Run as root in the dev compose because bind-mounted host files
+    user: "0:0"  # dev only — see security note above
-    # (./src, ./static) may not be readable by the image's non-root
-    # "timmy" user — this is the #1 cause of 403 errors on macOS.
-    # Production (docker-compose.prod.yml) uses no bind mounts and
-    # correctly runs as the Dockerfile's non-root USER.
-    user: "0:0"
     ports:
       - "8000:8000"
     volumes:
@@ -36,14 +53,13 @@ services:
       - ./static:/app/static  # live-reload: CSS/asset changes reflect immediately
     environment:
       DEBUG: "true"
-      # Point to host Ollama (Mac default).  Override in .env if different.
       OLLAMA_URL: "${OLLAMA_URL:-http://host.docker.internal:11434}"
       # Grok (xAI) — opt-in premium cloud backend
       GROK_ENABLED: "${GROK_ENABLED:-false}"
       XAI_API_KEY: "${XAI_API_KEY:-}"
       GROK_DEFAULT_MODEL: "${GROK_DEFAULT_MODEL:-grok-3-fast}"
     extra_hosts:
-      - "host.docker.internal:host-gateway"  # Linux compatibility
+      - "host.docker.internal:host-gateway"  # Linux: maps to host IP
     networks:
       - swarm-net
     restart: unless-stopped

docker/Dockerfile.test

@@ -0,0 +1,59 @@
+# ── Timmy Time — Test Runner Image ───────────────────────────────────────────
+#
+# Lean image with test dependencies baked in.  Designed to be used with
+# docker-compose.test.yml which bind-mounts src/, tests/, and static/
+# so you never rebuild for code changes — only when deps change.
+#
+# Build:  docker compose -f docker-compose.test.yml build
+# Run:    docker compose -f docker-compose.test.yml run --rm test
+#
+# The builder stage is shared with the production Dockerfile so
+# dependency layers stay cached across dev ↔ test ↔ prod builds.
+
+# ── Stage 1: Builder — export deps via Poetry, install via pip ──────────────
+FROM python:3.12-slim AS builder
+
+WORKDIR /build
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+        gcc curl \
+    && rm -rf /var/lib/apt/lists/*
+
+RUN pip install --no-cache-dir poetry poetry-plugin-export
+
+# Copy only dependency files (layer caching — rebuilds only when deps change)
+COPY pyproject.toml poetry.lock ./
+
+# Export ALL deps including dev/test extras
+RUN poetry export --extras swarm --extras telegram --extras dev \
+        --with dev --without-hashes \
+        -f requirements.txt -o requirements.txt
+
+RUN --mount=type=cache,target=/root/.cache/pip \
+    pip install --no-cache-dir -r requirements.txt
+
+
+# ── Stage 2: Test runtime ───────────────────────────────────────────────────
+FROM python:3.12-slim
+
+WORKDIR /app
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+        curl git \
+    && rm -rf /var/lib/apt/lists/*
+
+# Copy installed packages from builder
+COPY --from=builder /usr/local/lib/python3.12/site-packages \
+                    /usr/local/lib/python3.12/site-packages
+COPY --from=builder /usr/local/bin /usr/local/bin
+
+# Create directories for bind mounts
+RUN mkdir -p /app/src /app/tests /app/static /app/data
+
+ENV PYTHONPATH=/app/src:/app/tests
+ENV PYTHONUNBUFFERED=1
+ENV PYTHONDONTWRITEBYTECODE=1
+ENV TIMMY_TEST_MODE=1
+
+# Default: run pytest (overridable via docker-compose command)
+CMD ["pytest", "tests/", "-q", "--tb=short"]

docker/scripts/init-ollama.sh

@@ -2,11 +2,20 @@
 # ── Ollama Initialization Script ──────────────────────────────────────────────
 #
 # Starts Ollama and pulls models on first run.
+# Requires: curl (ships with the ollama image).
+# jq is installed at runtime if missing so we can parse /api/tags reliably
+# instead of fragile grep-based JSON extraction.
 
 set -e
 
 echo "🚀 Ollama startup — checking for models..."
 
+# ── Ensure jq is available (ollama image is Debian-based) ────────────────────
+if ! command -v jq &>/dev/null; then
+  echo "📦 Installing jq for reliable JSON parsing..."
+  apt-get update -qq && apt-get install -y -qq jq >/dev/null 2>&1 || true
+fi
+
 # Start Ollama in background
 ollama serve &
 OLLAMA_PID=$!
@@ -18,15 +27,26 @@ for i in {1..60}; do
     echo "✓ Ollama is ready"
     break
   fi
+  if [ "$i" -eq 60 ]; then
+    echo "❌ Ollama failed to start after 60 s"
+    exit 1
+  fi
   echo "  Attempt $i/60..."
   sleep 1
 done
 
-# Check if models are already present
+# Check if models are already present (jq with grep fallback)
 echo "📋 Checking available models..."
-MODELS=$(curl -s http://localhost:11434/api/tags | grep -o '"name":"[^"]*"' | wc -l)
+TAGS_JSON=$(curl -s http://localhost:11434/api/tags)
 
-if [ "$MODELS" -eq 0 ]; then
+if command -v jq &>/dev/null; then
+  MODELS=$(echo "$TAGS_JSON" | jq '.models | length')
+else
+  # Fallback: count "name" keys (less reliable but functional)
+  MODELS=$(echo "$TAGS_JSON" | grep -o '"name"' | wc -l)
+fi
+
+if [ "${MODELS:-0}" -eq 0 ]; then
   echo "📥 No models found. Pulling llama3.2..."
   ollama pull llama3.2 || echo "⚠️  Failed to pull llama3.2 (may already be pulling)"
 else

poetry.lock

@@ -458,7 +458,7 @@ files = [
     {file = "attrs-25.4.0-py3-none-any.whl", hash = "sha256:adcf7e2a1fb3b36ac48d97835bb6d8ade15b8dcce26aba8bf1d14847b57a3373"},
     {file = "attrs-25.4.0.tar.gz", hash = "sha256:16d5969b87f0859ef33a48b35d55ac1be6e42ae49d5e853b597db70c35c57e11"},
 ]
-markers = {main = "extra == \"discord\""}
+markers = {main = "extra == \"discord\" or extra == \"dev\""}
 
 [[package]]
 name = "audioop-lts"
@@ -538,8 +538,7 @@ version = "2.0.0"
 description = "Foreign Function Interface for Python calling C code."
 optional = false
 python-versions = ">=3.9"
-groups = ["dev"]
+groups = ["main", "dev"]
-markers = "os_name == \"nt\" and implementation_name != \"pypy\""
 files = [
     {file = "cffi-2.0.0-cp310-cp310-macosx_10_13_x86_64.whl", hash = "sha256:0cf2d91ecc3fcc0625c2c530fe004f82c110405f101548512cce44322fa8ac44"},
     {file = "cffi-2.0.0-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:f73b96c41e3b2adedc34a7356e64c8eb96e03a3782b535e043a986276ce12a49"},
@@ -626,6 +625,7 @@ files = [
     {file = "cffi-2.0.0-cp39-cp39-win_amd64.whl", hash = "sha256:b882b3df248017dba09d6b16defe9b5c407fe32fc7c65a9c69798e6175601be9"},
     {file = "cffi-2.0.0.tar.gz", hash = "sha256:44d1b5909021139fe36001ae048dbdde8214afa20200eda0f64c068cac5d5529"},
 ]
+markers = {main = "os_name == \"nt\" and implementation_name != \"pypy\" and extra == \"dev\"", dev = "os_name == \"nt\" and implementation_name != \"pypy\""}
 
 [package.dependencies]
 pycparser = {version = "*", markers = "implementation_name != \"PyPy\""}
@@ -800,7 +800,7 @@ version = "7.13.4"
 description = "Code coverage measurement for Python"
 optional = false
 python-versions = ">=3.10"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "coverage-7.13.4-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:0fc31c787a84f8cd6027eba44010517020e0d18487064cd3d8968941856d1415"},
     {file = "coverage-7.13.4-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:a32ebc02a1805adf637fc8dec324b5cdacd2e493515424f70ee33799573d661b"},
@@ -909,6 +909,7 @@ files = [
     {file = "coverage-7.13.4-py3-none-any.whl", hash = "sha256:1af1641e57cf7ba1bd67d677c9abdbcd6cc2ab7da3bca7fa1e2b7e50e65f2ad0"},
     {file = "coverage-7.13.4.tar.gz", hash = "sha256:e5c8f6ed1e61a8b2dcdf31eb0b9bbf0130750ca79c1c49eb898e2ad86f5ccc91"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [package.extras]
 toml = ["tomli ; python_full_version <= \"3.11.0a6\""]
@@ -1599,11 +1600,12 @@ version = "2.3.0"
 description = "brain-dead simple config-ini parsing"
 optional = false
 python-versions = ">=3.10"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "iniconfig-2.3.0-py3-none-any.whl", hash = "sha256:f631c04d2c48c52b84d0d0549c99ff3859c98df65b3101406327ecc7d53fbf12"},
     {file = "iniconfig-2.3.0.tar.gz", hash = "sha256:c76315c77db068650d49c5b56314774a7804df16fee4402c1f19d6d15d8c4730"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [[package]]
 name = "jinja2"
@@ -2490,11 +2492,12 @@ version = "1.3.0.post0"
 description = "Capture the outcome of Python function calls."
 optional = false
 python-versions = ">=3.7"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "outcome-1.3.0.post0-py2.py3-none-any.whl", hash = "sha256:e771c5ce06d1415e356078d3bdd68523f284b4ce5419828922b6871e65eda82b"},
     {file = "outcome-1.3.0.post0.tar.gz", hash = "sha256:9dcf02e65f2971b80047b377468e72a268e15c0af3cf1238e6ff14f7f91143b8"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [package.dependencies]
 attrs = ">=19.2.0"
@@ -2642,11 +2645,12 @@ version = "1.6.0"
 description = "plugin and hook calling mechanisms for python"
 optional = false
 python-versions = ">=3.9"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "pluggy-1.6.0-py3-none-any.whl", hash = "sha256:e920276dd6813095e9377c0bc5566d94c932c33b27a3e3945d8389c374dd4746"},
     {file = "pluggy-1.6.0.tar.gz", hash = "sha256:7dcc130b76258d33b90f61b658791dede3486c3e6bfb003ee5c9bfb396dd22f3"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [package.extras]
 dev = ["pre-commit", "tox"]
@@ -2842,12 +2846,12 @@ version = "3.0"
 description = "C parser in Python"
 optional = false
 python-versions = ">=3.10"
-groups = ["dev"]
+groups = ["main", "dev"]
-markers = "os_name == \"nt\" and implementation_name != \"pypy\" and implementation_name != \"PyPy\""
 files = [
     {file = "pycparser-3.0-py3-none-any.whl", hash = "sha256:b727414169a36b7d524c1c3e31839a521725078d7b2ff038656844266160a992"},
     {file = "pycparser-3.0.tar.gz", hash = "sha256:600f49d217304a5902ac3c37e1281c9fe94e4d0489de643a9504c5cdfdfc6b29"},
 ]
+markers = {main = "os_name == \"nt\" and implementation_name != \"pypy\" and implementation_name != \"PyPy\" and extra == \"dev\"", dev = "os_name == \"nt\" and implementation_name != \"pypy\" and implementation_name != \"PyPy\""}
 
 [[package]]
 name = "pydantic"
@@ -6617,12 +6621,13 @@ version = "1.7.1"
 description = "A Python SOCKS client module. See https://github.com/Anorov/PySocks for more information."
 optional = false
 python-versions = ">=2.7, !=3.0.*, !=3.1.*, !=3.2.*, !=3.3.*"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "PySocks-1.7.1-py27-none-any.whl", hash = "sha256:08e69f092cc6dbe92a0fdd16eeb9b9ffbc13cadfe5ca4c7bd92ffb078b293299"},
     {file = "PySocks-1.7.1-py3-none-any.whl", hash = "sha256:2725bd0a9925919b9b51739eea5f9e2bae91e83288108a9ad338b2e3a4435ee5"},
     {file = "PySocks-1.7.1.tar.gz", hash = "sha256:3f8804571ebe159c380ac6de37643bb4685970655d3bba243530d6558b799aa0"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [[package]]
 name = "pytest"
@@ -6630,11 +6635,12 @@ version = "9.0.2"
 description = "pytest: simple powerful testing with Python"
 optional = false
 python-versions = ">=3.10"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "pytest-9.0.2-py3-none-any.whl", hash = "sha256:711ffd45bf766d5264d487b917733b453d917afd2b0ad65223959f59089f875b"},
     {file = "pytest-9.0.2.tar.gz", hash = "sha256:75186651a92bd89611d1d9fc20f0b4345fd827c41ccd5c299a868a05d70edf11"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [package.dependencies]
 colorama = {version = ">=0.4", markers = "sys_platform == \"win32\""}
@@ -6652,11 +6658,12 @@ version = "1.3.0"
 description = "Pytest support for asyncio"
 optional = false
 python-versions = ">=3.10"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "pytest_asyncio-1.3.0-py3-none-any.whl", hash = "sha256:611e26147c7f77640e6d0a92a38ed17c3e9848063698d5c93d5aa7aa11cebff5"},
     {file = "pytest_asyncio-1.3.0.tar.gz", hash = "sha256:d7f52f36d231b80ee124cd216ffb19369aa168fc10095013c6b014a34d3ee9e5"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [package.dependencies]
 pytest = ">=8.2,<10"
@@ -6672,11 +6679,12 @@ version = "7.0.0"
 description = "Pytest plugin for measuring coverage."
 optional = false
 python-versions = ">=3.9"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "pytest_cov-7.0.0-py3-none-any.whl", hash = "sha256:3b8e9558b16cc1479da72058bdecf8073661c7f57f7d3c5f22a1c23507f2d861"},
     {file = "pytest_cov-7.0.0.tar.gz", hash = "sha256:33c97eda2e049a0c5298e91f519302a1334c26ac65c1a483d6206fd458361af1"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [package.dependencies]
 coverage = {version = ">=7.10.6", extras = ["toml"]}
@@ -6686,17 +6694,34 @@ pytest = ">=7"
 [package.extras]
 testing = ["process-tests", "pytest-xdist", "virtualenv"]
 
+[[package]]
+name = "pytest-randomly"
+version = "4.0.1"
+description = "Pytest plugin to randomly order tests and control random.seed."
+optional = false
+python-versions = ">=3.9"
+groups = ["main", "dev"]
+files = [
+    {file = "pytest_randomly-4.0.1-py3-none-any.whl", hash = "sha256:e0dfad2fd4f35e07beff1e47c17fbafcf98f9bf4531fd369d9260e2f858bfcb7"},
+    {file = "pytest_randomly-4.0.1.tar.gz", hash = "sha256:174e57bb12ac2c26f3578188490bd333f0e80620c3f47340158a86eca0593cd8"},
+]
+markers = {main = "extra == \"dev\""}
+
+[package.dependencies]
+pytest = "*"
+
 [[package]]
 name = "pytest-timeout"
 version = "2.4.0"
 description = "pytest plugin to abort hanging tests"
 optional = false
 python-versions = ">=3.7"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "pytest_timeout-2.4.0-py3-none-any.whl", hash = "sha256:c42667e5cdadb151aeb5b26d114aff6bdf5a907f176a007a30b940d3d865b5c2"},
     {file = "pytest_timeout-2.4.0.tar.gz", hash = "sha256:7e68e90b01f9eff71332b25001f85c75495fc4e3a836701876183c4bcfd0540a"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [package.dependencies]
 pytest = ">=7.0.0"
@@ -7213,11 +7238,12 @@ version = "4.41.0"
 description = "Official Python bindings for Selenium WebDriver"
 optional = false
 python-versions = ">=3.10"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "selenium-4.41.0-py3-none-any.whl", hash = "sha256:b8ccde8d2e7642221ca64af184a92c19eee6accf2e27f20f30472f5efae18eb1"},
     {file = "selenium-4.41.0.tar.gz", hash = "sha256:003e971f805231ad63e671783a2b91a299355d10cefb9de964c36ff3819115aa"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [package.dependencies]
 certifi = ">=2026.1.4"
@@ -7291,11 +7317,12 @@ version = "2.4.0"
 description = "Sorted Containers -- Sorted List, Sorted Dict, Sorted Set"
 optional = false
 python-versions = "*"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "sortedcontainers-2.4.0-py2.py3-none-any.whl", hash = "sha256:a163dcaede0f1c021485e957a39245190e74249897e2ae4b2aa38595db237ee0"},
     {file = "sortedcontainers-2.4.0.tar.gz", hash = "sha256:25caa5a06cc30b6b83d11423433f65d1f9d76c4c6a0c90e3379eaa43b9bfdb88"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [[package]]
 name = "sqlalchemy"
@@ -7626,11 +7653,12 @@ version = "0.33.0"
 description = "A friendly Python library for async concurrency and I/O"
 optional = false
 python-versions = ">=3.10"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "trio-0.33.0-py3-none-any.whl", hash = "sha256:3bd5d87f781d9b0192d592aef28691f8951d6c2e41b7e1da4c25cde6c180ae9b"},
     {file = "trio-0.33.0.tar.gz", hash = "sha256:a29b92b73f09d4b48ed249acd91073281a7f1063f09caba5dc70465b5c7aa970"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [package.dependencies]
 attrs = ">=23.2.0"
@@ -7646,11 +7674,12 @@ version = "0.12.2"
 description = "WebSocket library for Trio"
 optional = false
 python-versions = ">=3.8"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "trio_websocket-0.12.2-py3-none-any.whl", hash = "sha256:df605665f1db533f4a386c94525870851096a223adcb97f72a07e8b4beba45b6"},
     {file = "trio_websocket-0.12.2.tar.gz", hash = "sha256:22c72c436f3d1e264d0910a3951934798dcc5b00ae56fc4ee079d46c7cf20fae"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [package.dependencies]
 outcome = ">=1.2.0"
@@ -7983,11 +8012,12 @@ version = "1.9.0"
 description = "WebSocket client for Python with low level API options"
 optional = false
 python-versions = ">=3.9"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "websocket_client-1.9.0-py3-none-any.whl", hash = "sha256:af248a825037ef591efbf6ed20cc5faa03d3b47b9e5a2230a529eeee1c1fc3ef"},
     {file = "websocket_client-1.9.0.tar.gz", hash = "sha256:9e813624b6eb619999a97dc7958469217c3176312b3a16a4bd1bc7e08a46ec98"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [package.extras]
 docs = ["Sphinx (>=6.0)", "myst-parser (>=2.0.0)", "sphinx_rtd_theme (>=1.1.0)"]
@@ -8071,11 +8101,12 @@ version = "1.3.2"
 description = "Pure-Python WebSocket protocol implementation"
 optional = false
 python-versions = ">=3.10"
-groups = ["dev"]
+groups = ["main", "dev"]
 files = [
     {file = "wsproto-1.3.2-py3-none-any.whl", hash = "sha256:61eea322cdf56e8cc904bd3ad7573359a242ba65688716b0710a5eb12beab584"},
     {file = "wsproto-1.3.2.tar.gz", hash = "sha256:b86885dcf294e15204919950f666e06ffc6c7c114ca900b060d6e16293528294"},
 ]
+markers = {main = "extra == \"dev\""}
 
 [package.dependencies]
 h11 = ">=0.16.0,<1"
@@ -8228,6 +8259,7 @@ propcache = ">=0.2.1"
 
 [extras]
 bigbrain = ["airllm"]
+dev = ["pytest", "pytest-asyncio", "pytest-cov", "pytest-randomly", "pytest-timeout", "selenium"]
 discord = ["discord.py"]
 swarm = ["redis"]
 telegram = ["python-telegram-bot"]
@@ -8236,4 +8268,4 @@ voice = ["pyttsx3"]
 [metadata]
 lock-version = "2.1"
 python-versions = ">=3.11,<4"
-content-hash = "546e3cc56929a6b988223fbc685fdb61468fbe5a50249be624742edca30f137e"
+content-hash = "8e608d71fafb99eda990a90f7879127522ec03fcd2bd34b115d2b4fde4c0fe87"

pyproject.toml

@@ -54,6 +54,7 @@ pytest-asyncio = { version = ">=0.24.0", optional = true }
 pytest-cov = { version = ">=5.0.0", optional = true }
 pytest-timeout = { version = ">=2.3.0", optional = true }
 selenium = { version = ">=4.20.0", optional = true }
+pytest-randomly = { version = ">=3.16.0", optional = true }
 
 [tool.poetry.extras]
 swarm = ["redis"]
@@ -61,7 +62,7 @@ telegram = ["python-telegram-bot"]
 discord = ["discord.py"]
 bigbrain = ["airllm"]
 voice = ["pyttsx3"]
-dev = ["pytest", "pytest-asyncio", "pytest-cov", "pytest-timeout", "selenium"]
+dev = ["pytest", "pytest-asyncio", "pytest-cov", "pytest-timeout", "pytest-randomly", "selenium"]
 
 [tool.poetry.group.dev.dependencies]
 pytest = ">=8.0.0"
@@ -69,6 +70,7 @@ pytest-asyncio = ">=0.24.0"
 pytest-cov = ">=5.0.0"
 pytest-timeout = ">=2.3.0"
 selenium = ">=4.20.0"
+pytest-randomly = "^4.0.1"
 
 [tool.poetry.scripts]
 timmy = "timmy.cli:main"
@@ -85,9 +87,15 @@ addopts = "-v --tb=short --timeout=30"
 markers = [
     "unit: Unit tests (fast, no I/O)",
     "integration: Integration tests (may use SQLite)",
+    "functional: Functional tests (real HTTP requests, no mocking)",
+    "e2e: End-to-end tests (full system, may be slow)",
     "dashboard: Dashboard route tests",
     "swarm: Swarm coordinator tests",
     "slow: Tests that take >1 second",
+    "selenium: Requires Selenium and Chrome (browser automation)",
+    "docker: Requires Docker and docker-compose",
+    "ollama: Requires Ollama service running",
+    "skip_ci: Skip in CI environment (local development only)",
 ]
 
 [tool.coverage.run]

tests/conftest.py

@@ -186,6 +186,32 @@ def db_connection():
 
 # ── Additional Clean Test Fixtures ──────────────────────────────────────────
 
+@pytest.fixture(autouse=True)
+def tmp_swarm_db(tmp_path, monkeypatch):
+    """Point all swarm SQLite paths to a temp directory for test isolation.
+
+    This is the single source of truth — individual test files should NOT
+    redefine this fixture.  All eight swarm modules that carry a module-level
+    DB_PATH are patched here so every test gets a clean, ephemeral database.
+    """
+    db_path = tmp_path / "swarm.db"
+    for module in [
+        "swarm.tasks",
+        "swarm.registry",
+        "swarm.stats",
+        "swarm.learner",
+        "swarm.routing",
+        "swarm.event_log",
+        "swarm.task_queue.models",
+        "swarm.work_orders.models",
+    ]:
+        try:
+            monkeypatch.setattr(f"{module}.DB_PATH", db_path)
+        except AttributeError:
+            pass  # Module may not be importable in minimal test envs
+    yield db_path
+
+
 @pytest.fixture
 def mock_ollama_client():
     """Provide a mock Ollama client for unit tests."""

tests/dashboard/test_dashboard_routes.py

@@ -7,16 +7,6 @@ import pytest
 from fastapi.testclient import TestClient
 
 
-@pytest.fixture(autouse=True)
-def tmp_swarm_db(tmp_path, monkeypatch):
-    """Point swarm SQLite to a temp directory for test isolation."""
-    db_path = tmp_path / "swarm.db"
-    monkeypatch.setattr("swarm.tasks.DB_PATH", db_path)
-    monkeypatch.setattr("swarm.registry.DB_PATH", db_path)
-    monkeypatch.setattr("swarm.stats.DB_PATH", db_path)
-    yield db_path
-
-
 @pytest.fixture
 def client():
     from dashboard.app import app

tests/e2e/test_docker_deployment.py

@@ -1,175 +1,120 @@
 """End-to-end tests for Docker deployment.
 
-These tests verify that the Dockerized application starts correctly,
+These tests verify that Dockerfiles and compose configs are present,
-responds to requests, and all services are properly orchestrated.
+syntactically valid, and declare the expected services and settings.
 """
 
-import pytest
-import subprocess
-import time
-import requests
 import json
+import subprocess
+
+import pytest
 from pathlib import Path
 
-
+PROJECT_ROOT = Path(__file__).parent.parent.parent
-@pytest.fixture(scope="module")
-def docker_compose_file():
-    """Return the path to the docker-compose file."""
-    return Path(__file__).parent.parent.parent / "docker-compose.enhanced.yml"
-
-
-@pytest.fixture(scope="module")
-def docker_services_running(docker_compose_file):
-    """Start Docker services for testing."""
-    if not docker_compose_file.exists():
-        pytest.skip("docker-compose.enhanced.yml not found")
-    
-    # Start services
-    result = subprocess.run(
-        ["docker", "compose", "-f", str(docker_compose_file), "up", "-d"],
-        capture_output=True,
-        text=True,
-    )
-    
-    if result.returncode != 0:
-        pytest.skip(f"Failed to start Docker services: {result.stderr}")
-    
-    # Wait for services to be ready
-    time.sleep(10)
-    
-    yield
-    
-    # Cleanup
-    subprocess.run(
-        ["docker", "compose", "-f", str(docker_compose_file), "down"],
-        capture_output=True,
-    )
 
 
 @pytest.mark.skipif(
     subprocess.run(["which", "docker"], capture_output=True).returncode != 0,
-    reason="Docker not installed"
+    reason="Docker not installed",
 )
-def test_docker_compose_file_exists():
+class TestDockerComposeFiles:
-    """Test that docker-compose.enhanced.yml exists."""
+    """Validate that all compose files exist and parse cleanly."""
-    compose_file = Path(__file__).parent.parent.parent / "docker-compose.enhanced.yml"
+
-    assert compose_file.exists(), "docker-compose.enhanced.yml should exist"
+    def test_base_compose_exists(self):
+        assert (PROJECT_ROOT / "docker-compose.yml").exists()
+
+    def test_dev_overlay_exists(self):
+        assert (PROJECT_ROOT / "docker-compose.dev.yml").exists()
+
+    def test_prod_compose_exists(self):
+        assert (PROJECT_ROOT / "docker-compose.prod.yml").exists()
+
+    def test_test_compose_exists(self):
+        assert (PROJECT_ROOT / "docker-compose.test.yml").exists()
+
+    def test_microservices_compose_exists(self):
+        assert (PROJECT_ROOT / "docker-compose.microservices.yml").exists()
+
+    def test_base_compose_syntax(self):
+        result = subprocess.run(
+            ["docker", "compose", "-f", str(PROJECT_ROOT / "docker-compose.yml"), "config"],
+            capture_output=True,
+            text=True,
+        )
+        assert result.returncode == 0, f"Docker Compose syntax error: {result.stderr}"
+
+    def test_microservices_compose_services_defined(self):
+        result = subprocess.run(
+            [
+                "docker", "compose",
+                "-f", str(PROJECT_ROOT / "docker-compose.microservices.yml"),
+                "config", "--format", "json",
+            ],
+            capture_output=True,
+            text=True,
+        )
+        assert result.returncode == 0, f"Config error: {result.stderr}"
+        config = json.loads(result.stdout)
+        services = config.get("services", {})
+        assert "ollama" in services, "ollama service should be defined"
+        assert "dashboard" in services, "dashboard service should be defined"
+        assert "timmy" in services, "timmy service should be defined"
+
+    def test_microservices_compose_content(self):
+        content = (PROJECT_ROOT / "docker-compose.microservices.yml").read_text()
+        assert "ollama" in content
+        assert "dashboard" in content
+        assert "timmy" in content
+        assert "timmy-net" in content
+        assert "ollama-data" in content
+        assert "timmy-data" in content
+
+    def test_test_compose_has_test_runner(self):
+        content = (PROJECT_ROOT / "docker-compose.test.yml").read_text()
+        assert "test:" in content, "Test compose should define a 'test' service"
+        assert "TIMMY_TEST_MODE" in content
+        assert "pytest" in content
 
 
-@pytest.mark.skipif(
+class TestDockerfiles:
-    subprocess.run(["which", "docker"], capture_output=True).returncode != 0,
+    """Validate the primary Dockerfile and specialised images."""
-    reason="Docker not installed"
+
-)
+    def test_dockerfile_exists(self):
-def test_docker_compose_syntax():
+        assert (PROJECT_ROOT / "Dockerfile").exists()
-    """Test that docker-compose file has valid syntax."""
+
-    compose_file = Path(__file__).parent.parent.parent / "docker-compose.enhanced.yml"
+    def test_dockerfile_ollama_exists(self):
-    
+        assert (PROJECT_ROOT / "docker" / "Dockerfile.ollama").exists()
-    result = subprocess.run(
+
-        ["docker", "compose", "-f", str(compose_file), "config"],
+    def test_dockerfile_agent_exists(self):
-        capture_output=True,
+        assert (PROJECT_ROOT / "docker" / "Dockerfile.agent").exists()
-        text=True,
+
+    def test_dockerfile_dashboard_exists(self):
+        assert (PROJECT_ROOT / "docker" / "Dockerfile.dashboard").exists()
+
+    def test_dockerfile_test_exists(self):
+        assert (PROJECT_ROOT / "docker" / "Dockerfile.test").exists()
+
+    def test_dockerfile_health_check(self):
+        content = (PROJECT_ROOT / "Dockerfile").read_text()
+        assert "HEALTHCHECK" in content, "Dockerfile should include HEALTHCHECK"
+        assert "/health" in content
+
+    def test_dockerfile_non_root_user(self):
+        content = (PROJECT_ROOT / "Dockerfile").read_text()
+        assert "USER timmy" in content
+        assert "groupadd -r timmy" in content
+
+    @pytest.mark.skipif(
+        subprocess.run(["which", "docker"], capture_output=True).returncode != 0,
+        reason="Docker not installed",
     )
-    
+    def test_docker_image_build(self):
-    assert result.returncode == 0, f"Docker Compose syntax error: {result.stderr}"
+        result = subprocess.run(
-
+            ["docker", "build", "-t", "timmy-time:test", "."],
-
+            cwd=PROJECT_ROOT,
-@pytest.mark.skipif(
+            capture_output=True,
-    subprocess.run(["which", "docker"], capture_output=True).returncode != 0,
+            text=True,
-    reason="Docker not installed"
+            timeout=300,
-)
+        )
-def test_dockerfile_exists():
+        if result.returncode != 0:
-    """Test that Dockerfile exists."""
+            pytest.skip(f"Docker build failed: {result.stderr}")
-    dockerfile = Path(__file__).parent.parent.parent / "Dockerfile"
-    assert dockerfile.exists(), "Dockerfile should exist"
-
-
-@pytest.mark.skipif(
-    subprocess.run(["which", "docker"], capture_output=True).returncode != 0,
-    reason="Docker not installed"
-)
-def test_dockerfile_ollama_exists():
-    """Test that Dockerfile.ollama exists."""
-    dockerfile = Path(__file__).parent.parent.parent / "Dockerfile.ollama"
-    assert dockerfile.exists(), "Dockerfile.ollama should exist"
-
-
-@pytest.mark.skipif(
-    subprocess.run(["which", "docker"], capture_output=True).returncode != 0,
-    reason="Docker not installed"
-)
-def test_docker_image_build():
-    """Test that the Docker image can be built."""
-    result = subprocess.run(
-        ["docker", "build", "-t", "timmy-time:test", "."],
-        cwd=Path(__file__).parent.parent.parent,
-        capture_output=True,
-        text=True,
-        timeout=300,
-    )
-    
-    # Don't fail if build fails, just skip
-    if result.returncode != 0:
-        pytest.skip(f"Docker build failed: {result.stderr}")
-
-
-@pytest.mark.skipif(
-    subprocess.run(["which", "docker"], capture_output=True, shell=True).returncode != 0,
-    reason="Docker not installed"
-)
-def test_docker_compose_services_defined():
-    """Test that docker-compose defines all required services."""
-    compose_file = Path(__file__).parent.parent.parent / "docker-compose.enhanced.yml"
-    
-    result = subprocess.run(
-        ["docker", "compose", "-f", str(compose_file), "config"],
-        capture_output=True,
-        text=True,
-    )
-    
-    assert result.returncode == 0, "Docker Compose config should be valid"
-    
-    config = json.loads(result.stdout)
-    services = config.get("services", {})
-    
-    # Check for required services
-    assert "ollama" in services, "ollama service should be defined"
-    assert "dashboard" in services, "dashboard service should be defined"
-    assert "timmy" in services, "timmy service should be defined"
-
-
-def test_docker_compose_enhanced_yml_content():
-    """Test that docker-compose.enhanced.yml has correct configuration."""
-    compose_file = Path(__file__).parent.parent.parent / "docker-compose.enhanced.yml"
-    
-    with open(compose_file) as f:
-        content = f.read()
-    
-    # Check for key configurations
-    assert "ollama" in content, "Should reference ollama service"
-    assert "dashboard" in content, "Should reference dashboard service"
-    assert "timmy" in content, "Should reference timmy agent"
-    assert "swarm-net" in content, "Should define swarm network"
-    assert "ollama-data" in content, "Should define ollama-data volume"
-    assert "timmy-data" in content, "Should define timmy-data volume"
-
-
-def test_dockerfile_health_check():
-    """Test that Dockerfile includes health check."""
-    dockerfile = Path(__file__).parent.parent.parent / "Dockerfile"
-    
-    with open(dockerfile) as f:
-        content = f.read()
-    
-    assert "HEALTHCHECK" in content, "Dockerfile should include HEALTHCHECK"
-    assert "/health" in content, "Health check should use /health endpoint"
-
-
-def test_dockerfile_non_root_user():
-    """Test that Dockerfile runs as non-root user."""
-    dockerfile = Path(__file__).parent.parent.parent / "Dockerfile"
-    
-    with open(dockerfile) as f:
-        content = f.read()
-    
-    assert "USER timmy" in content, "Dockerfile should run as non-root user"
-    assert "groupadd -r timmy" in content, "Dockerfile should create timmy user"

tests/swarm/test_coordinator.py

@@ -8,17 +8,6 @@ import pytest
 from unittest.mock import AsyncMock, patch
 
 
-@pytest.fixture(autouse=True)
-def tmp_swarm_db(tmp_path, monkeypatch):
-    """Point swarm SQLite to a temp directory for test isolation."""
-    db_path = tmp_path / "swarm.db"
-    monkeypatch.setattr("swarm.tasks.DB_PATH", db_path)
-    monkeypatch.setattr("swarm.registry.DB_PATH", db_path)
-    monkeypatch.setattr("swarm.stats.DB_PATH", db_path)
-    monkeypatch.setattr("swarm.learner.DB_PATH", db_path)
-    yield db_path
-
-
 # ── Coordinator: Agent lifecycle ─────────────────────────────────────────────
 
 def test_coordinator_spawn_agent():

tests/swarm/test_swarm.py

@@ -10,14 +10,6 @@ import pytest
 
 # ── Tasks CRUD ───────────────────────────────────────────────────────────────
 
-@pytest.fixture(autouse=True)
-def tmp_swarm_db(tmp_path, monkeypatch):
-    """Point swarm SQLite to a temp directory for test isolation."""
-    db_path = tmp_path / "swarm.db"
-    monkeypatch.setattr("swarm.tasks.DB_PATH", db_path)
-    monkeypatch.setattr("swarm.registry.DB_PATH", db_path)
-    yield db_path
-
 
 def test_create_task():
     from swarm.tasks import create_task

tests/swarm/test_swarm_node.py

@@ -7,15 +7,6 @@ import pytest
 from unittest.mock import MagicMock, patch
 
 
-@pytest.fixture(autouse=True)
-def tmp_swarm_db(tmp_path, monkeypatch):
-    """Point swarm SQLite to a temp directory for test isolation."""
-    db_path = tmp_path / "swarm.db"
-    monkeypatch.setattr("swarm.tasks.DB_PATH", db_path)
-    monkeypatch.setattr("swarm.registry.DB_PATH", db_path)
-    yield db_path
-
-
 def _make_node(agent_id="node-1", name="TestNode"):
     from swarm.comms import SwarmComms
     from swarm.swarm_node import SwarmNode

tests/swarm/test_swarm_personas.py

@@ -4,18 +4,6 @@ import pytest
 from unittest.mock import MagicMock
 
 
-# ── Fixture: redirect SQLite DB to a temp directory ──────────────────────────
-
-@pytest.fixture(autouse=True)
-def tmp_swarm_db(tmp_path, monkeypatch):
-    db_path = tmp_path / "swarm.db"
-    monkeypatch.setattr("swarm.tasks.DB_PATH", db_path)
-    monkeypatch.setattr("swarm.registry.DB_PATH", db_path)
-    monkeypatch.setattr("swarm.stats.DB_PATH", db_path)
-    monkeypatch.setattr("swarm.learner.DB_PATH", db_path)
-    yield db_path
-
-
 # ── personas.py ───────────────────────────────────────────────────────────────
 
 def test_all_nine_personas_defined():

tests/swarm/test_swarm_recovery.py

@@ -3,16 +3,6 @@
 import pytest
 
 
-@pytest.fixture(autouse=True)
-def tmp_swarm_db(tmp_path, monkeypatch):
-    """Isolate SQLite writes to a temp directory."""
-    db = tmp_path / "swarm.db"
-    monkeypatch.setattr("swarm.tasks.DB_PATH", db)
-    monkeypatch.setattr("swarm.registry.DB_PATH", db)
-    monkeypatch.setattr("swarm.stats.DB_PATH", db)
-    yield db
-
-
 # ── reconcile_on_startup: return shape ───────────────────────────────────────
 
 def test_reconcile_returns_summary_keys():

tests/swarm/test_swarm_stats.py

@@ -3,13 +3,6 @@
 import pytest
 
 
-@pytest.fixture(autouse=True)
-def tmp_swarm_db(tmp_path, monkeypatch):
-    db_path = tmp_path / "swarm.db"
-    monkeypatch.setattr("swarm.stats.DB_PATH", db_path)
-    yield db_path
-
-
 # ── record_bid ────────────────────────────────────────────────────────────────
 
 def test_record_bid_returns_id():

tests/timmy/test_agent_runner.py

@@ -11,14 +11,6 @@ from unittest.mock import AsyncMock, MagicMock, patch
 import pytest
 
 
-@pytest.fixture(autouse=True)
-def tmp_swarm_db(tmp_path, monkeypatch):
-    db_path = tmp_path / "swarm.db"
-    monkeypatch.setattr("swarm.tasks.DB_PATH", db_path)
-    monkeypatch.setattr("swarm.registry.DB_PATH", db_path)
-    yield db_path
-
-
 def test_agent_runner_module_is_importable():
     """The agent_runner module should import without errors."""
     import swarm.agent_runner