feat(hermes-workspace): add combined entrypoint script with gateway health check and graceful shutdown

ai/compose.yml

@@ -32,7 +32,7 @@ services:
         - default
     container_name: hermes
     entrypoint: ["/bin/bash", "-c",
-      "bash /opt/data/hermes-tools/install.sh && . /opt/hermes/.venv/bin/activate && uv pip install openai 'mautrix[encryption]' -q && exec /usr/bin/tini -g -- /opt/hermes/docker/entrypoint.sh \"$@\"",
+      "bash /opt/data/hermes-tools/install.sh && exec /usr/bin/tini -g -- /opt/hermes/docker/entrypoint.sh \"$@\"",
       "hermes-entrypoint"]
     restart: always
     # Gateway run enables the internal API server on port 8642
@@ -58,8 +58,6 @@ services:
       - /mnt/HoardingCow_docker_data/Syncthing/telos-ro:/opt/data/telos-ro:ro
       # Syncthing-shared inbox — write tasks here, they sync to user's laptop
       - /mnt/HoardingCow_docker_data/Syncthing/telos-rw:/opt/data/telos-rw:rw
-      # Persist Python venv across container recreation (Matrix bridge deps, etc.)
-      - /mnt/HoardingCow_docker_data/Hermes/venv:/opt/hermes/.venv
     devices:
       - /dev/kfd:/dev/kfd
       - /dev/dri:/dev/dri
@@ -68,8 +66,6 @@ services:
       - "26"
     networks:
       - ai_backend
-    depends_on:
-      - honcho
 
   syncthing:
     image: syncthing/syncthing:latest
@@ -133,66 +129,6 @@ services:
       - "303"
       - "26"
 
-  # --- Honcho: AI-native user modeling ---
-  honcho:
-    build: ./honcho
-    container_name: honcho
-    restart: unless-stopped
-    ports:
-      - "127.0.0.1:8000:8000"
-    environment:
-      - DB_CONNECTION_URI=postgresql+psycopg://honcho:honcho_pass@honcho-db:5432/honcho
-      - CACHE_URL=redis://honcho-redis:6379/0
-      - CACHE_ENABLED=true
-    volumes:
-      - /mnt/HoardingCow_docker_data/Honcho/data:/app/data
-    networks:
-      - ai_backend
-    depends_on:
-      honcho-db:
-        condition: service_healthy
-      honcho-redis:
-        condition: service_healthy
-
-  honcho-db:
-    image: pgvector/pgvector:pg15
-    container_name: honcho-db
-    restart: unless-stopped
-    ports:
-      - "127.0.0.1:5432:5432"
-    command: ["postgres", "-c", "max_connections=200"]
-    environment:
-      - POSTGRES_DB=honcho
-      - POSTGRES_USER=honcho
-      - POSTGRES_PASSWORD=honcho_pass
-      - PGDATA=/var/lib/postgresql/data/pgdata
-    volumes:
-      - /mnt/HoardingCow_docker_data/Honcho/postgres:/var/lib/postgresql/data
-      - ./honcho/init-db.sql:/docker-entrypoint-initdb.d/init.sql:ro
-    networks:
-      - ai_backend
-    healthcheck:
-      test: ["CMD-SHELL", "pg_isready -U honcho -d honcho"]
-      interval: 5s
-      timeout: 5s
-      retries: 5
-
-  honcho-redis:
-    image: redis:8
-    container_name: honcho-redis
-    restart: unless-stopped
-    ports:
-      - "127.0.0.1:6379:6379"
-    volumes:
-      - /mnt/HoardingCow_docker_data/Honcho/redis:/data
-    networks:
-      - ai_backend
-    healthcheck:
-      test: ["CMD-SHELL", "redis-cli ping"]
-      interval: 5s
-      timeout: 5s
-      retries: 5
-
 networks:
   ai_net:
     external: true

ai/hermes-workspace/entrypoint-combined.sh

@@ -0,0 +1,154 @@
+#!/bin/bash
+# ── Hermes Workspace Combined Entrypoint ──
+# Waits for the Hermes gateway container (hermes:8642) to become healthy,
+# then starts the Hermes Workspace web UI in the foreground.
+# Supports graceful shutdown via SIGTERM/SIGINT.
+# ──────────────────────────────────────────
+
+set -euo pipefail
+
+# ── Configuration ──────────────────────────────────────────────
+GATEWAY_HOST="${GATEWAY_HOST:-hermes}"
+GATEWAY_PORT="${GATEWAY_PORT:-8642}"
+GATEWAY_URL="http://${GATEWAY_HOST}:${GATEWAY_PORT}"
+
+HEALTH_ENDPOINT="${HEALTH_ENDPOINT:-/health}"
+MAX_RETRIES="${HEALTH_MAX_RETRIES:-60}"
+RETRY_INTERVAL="${HEALTH_RETRY_INTERVAL:-2}"
+
+WORKSPACE_DIR="${WORKSPACE_DIR:-/workspace}"
+WORKSPACE_ENTRY="${WORKSPACE_ENTRY:-server-entry.js}"
+
+PID_FILE="${PID_FILE:-/tmp/workspace.pid}"
+
+# ── Logging ────────────────────────────────────────────────────
+log_info()  { echo "[$(date '+%Y-%m-%d %H:%M:%S')] [INFO]  $*"; }
+log_warn()  { echo "[$(date '+%Y-%m-%d %H:%M:%S')] [WARN]  $*"; }
+log_error() { echo "[$(date '+%Y-%m-%d %H:%M:%S')] [ERROR] $*"; }
+
+# ── Graceful Shutdown ──────────────────────────────────────────
+_workspace_pid=""
+_shutting_down=false
+
+cleanup() {
+    if [ "$_shutting_down" = true ]; then
+        return
+    fi
+    _shutting_down=true
+
+    log_info "Shutdown signal received, cleaning up..."
+
+    # Stop workspace process if running
+    if [ -n "$_workspace_pid" ] && kill -0 "$_workspace_pid" 2>/dev/null; then
+        log_info "Stopping workspace (PID: $_workspace_pid)..."
+        kill -TERM "$_workspace_pid" 2>/dev/null || true
+
+        # Give it time to shut down gracefully
+        local wait_sec=10
+        while kill -0 "$_workspace_pid" 2>/dev/null && [ "$wait_sec" -gt 0 ]; do
+            sleep 1
+            wait_sec=$((wait_sec - 1))
+        done
+
+        # Force kill if still running
+        if kill -0 "$_workspace_pid" 2>/dev/null; then
+            log_warn "Workspace did not shut down gracefully, force killing..."
+            kill -KILL "$_workspace_pid" 2>/dev/null || true
+        fi
+    fi
+
+    # Clean up PID file
+    [ -f "$PID_FILE" ] && rm -f "$PID_FILE"
+
+    log_info "Shutdown complete."
+    exit 0
+}
+
+# Trap termination signals for graceful shutdown
+trap cleanup SIGTERM SIGINT
+
+# ── Gateway Health Check ───────────────────────────────────────
+wait_for_gateway() {
+    local url="${GATEWAY_URL}${HEALTH_ENDPOINT}"
+    local retries="$MAX_RETRIES"
+    local interval="$RETRY_INTERVAL"
+    local attempt=0
+
+    log_info "Waiting for Hermes gateway at ${GATEWAY_URL}..."
+    log_info "Max retries: ${retries}, interval: ${interval}s"
+
+    while [ "$attempt" -lt "$retries" ]; do
+        attempt=$((attempt + 1))
+
+        if curl -fsS "${url}" >/dev/null 2>&1; then
+            log_info "Gateway is healthy after ${attempt} attempt(s) (${GATEWAY_URL})"
+            return 0
+        fi
+
+        if [ "$attempt" -lt "$retries" ]; then
+            log_info "Gateway not ready yet (attempt ${attempt}/${retries}), retrying in ${interval}s..."
+            sleep "$interval"
+        fi
+    done
+
+    log_error "Gateway did not become healthy after ${retries} attempts ($((retries * interval))s)"
+    return 1
+}
+
+# ── Workspace Startup ──────────────────────────────────────────
+start_workspace() {
+    local entry="${WORKSPACE_DIR}/${WORKSPACE_ENTRY}"
+
+    if [ ! -d "$WORKSPACE_DIR" ]; then
+        log_error "Workspace directory not found: ${WORKSPACE_DIR}"
+        return 1
+    fi
+
+    if [ ! -f "$entry" ]; then
+        log_error "Workspace entry point not found: ${entry}"
+        return 1
+    fi
+
+    log_info "Starting Hermes Workspace web UI..."
+    log_info "  Directory: ${WORKSPACE_DIR}"
+    log_info "  Entry:     ${entry}"
+
+    cd "$WORKSPACE_DIR"
+
+    # Start workspace in background so we can trap signals
+    node --max-old-space-size=2048 "${entry}" &
+    _workspace_pid=$!
+    echo "$_workspace_pid" > "$PID_FILE"
+
+    log_info "Workspace started (PID: ${_workspace_pid})"
+
+    # Wait for workspace process
+    wait "$_workspace_pid"
+    local exit_code=$?
+
+    log_info "Workspace exited with code ${exit_code}"
+    return "$exit_code"
+}
+
+# ── Main ───────────────────────────────────────────────────────
+main() {
+    log_info "=== Hermes Workspace Combined Entrypoint ==="
+    log_info "Gateway:   ${GATEWAY_URL}"
+    log_info "Workspace: ${WORKSPACE_DIR}/${WORKSPACE_ENTRY}"
+    log_info "PID file:  ${PID_FILE}"
+
+    # Wait for gateway to be healthy
+    if ! wait_for_gateway; then
+        log_warn "Proceeding without confirmed gateway health..."
+    fi
+
+    # Start the workspace
+    start_workspace
+    local exit_code=$?
+
+    log_info "Entrypoint exiting with code ${exit_code}"
+    return "$exit_code"
+}
+
+# Run main; exit with its return code
+main "$@"

ai/hermes/Dockerfile

@@ -48,14 +48,6 @@ RUN apt-get update && \
 # ---------- UV ----------
 COPY --chmod=0755 --from=ghcr.io/astral-sh/uv:latest /uv /usr/local/bin/
 
-WORKDIR /opt/hermes
-
-# ---------- Memory provider dependencies ----------
-# httpx: HTTP client for OpenViking plugin
-# honcho-ai: already installed in upstream image (v2.1.1+)
-RUN . /opt/hermes/.venv/bin/activate && \
-    uv pip install --no-cache-dir httpx
-
 # ---------- Piper TTS ----------
 RUN . /opt/hermes/.venv/bin/activate && \
     uv pip install --no-cache-dir piper-tts sounddevice numpy && \

ai/honcho/Dockerfile

@@ -1,42 +0,0 @@
-# build stage — fetches and builds Honcho from source
-# Using buildkit cache mounts for speed across rebuilds
-FROM python:3.13-slim-bookworm AS builder
-
-RUN apt-get update && \
-    apt-get install -y --no-install-recommends git && \
-    rm -rf /var/lib/apt/lists/*
-
-COPY --from=ghcr.io/astral-sh/uv:0.9.24 /uv /bin/uv
-
-# Clone Honcho at a pinned commit for reproducibility
-ARG HONCHO_REPO=https://github.com/plastic-labs/honcho
-ARG HONCHO_REF=main
-RUN git clone --depth 1 --branch ${HONCHO_REF} ${HONCHO_REPO} /app
-
-WORKDIR /app
-
-ENV UV_COMPILE_BYTECODE=1
-ENV UV_LINK_MODE=copy
-
-RUN --mount=type=cache,target=/root/.cache/uv \
-    uv sync --frozen --no-group dev
-
-# --- runtime stage ---
-FROM python:3.13-slim-bookworm
-
-RUN groupadd --system app && \
-    useradd --system --gid app --create-home app
-
-COPY --from=builder /app /app
-COPY --from=builder /root/.cache/uv /root/.cache/uv
-
-WORKDIR /app
-ENV PATH="/app/.venv/bin:$PATH"
-ENV HOME=/app
-
-COPY --chown=app:app config.toml /app/config.toml
-
-USER app
-EXPOSE 8000
-
-CMD ["fastapi", "run", "--host", "0.0.0.0", "src/main.py"]

ai/honcho/config.toml

@@ -1,117 +0,0 @@
-[app]
-LOG_LEVEL = "INFO"
-MAX_MESSAGE_SIZE = 25000
-EMBED_MESSAGES = true
-NAMESPACE = "honcho"
-
-[db]
-CONNECTION_URI = "postgresql+psycopg://honcho:honcho_pass@honcho-db:5432/honcho"
-SCHEMA = "public"
-POOL_SIZE = 10
-MAX_OVERFLOW = 20
-
-[auth]
-USE_AUTH = false
-
-[sentry]
-ENABLED = false
-
-[telemetry]
-ENABLED = false
-
-[webhook]
-ENABLED = false
-
-[cache]
-ENABLED = true
-URL = "redis://honcho-redis:6379/0"
-
-[llm]
-DEFAULT_MAX_TOKENS = 4096
-
-# Embeddings via Ollama (nomic-embed-text recommended on this system)
-[embedding]
-VECTOR_DIMENSIONS = 768
-MAX_INPUT_TOKENS = 8192
-
-[embedding.model_config]
-transport = "openai"
-model = "nomic-embed-text"
-base_url = "http://ollama:11434/v1"
-
-# --- Deriver (user representation builder) ---
-[deriver]
-ENABLED = true
-WORKERS = 1
-POLLING_SLEEP_INTERVAL_SECONDS = 5.0
-FLUSH_ENABLED = true
-
-[deriver.model_config]
-transport = "openai"
-model = "hermes-3"
-base_url = "http://ollama:11434/v1"
-
-# --- Dialectic ---
-[dialectic]
-MAX_INPUT_TOKENS = 4096
-SESSION_HISTORY_MAX_TOKENS = 8192
-
-[dialectic.levels.minimal]
-MAX_TOOL_ITERATIONS = 1
-MAX_OUTPUT_TOKENS = 512
-[dialectic.levels.minimal.model_config]
-transport = "openai"
-model = "hermes-3"
-base_url = "http://ollama:11434/v1"
-
-[dialectic.levels.low]
-MAX_TOOL_ITERATIONS = 3
-[dialectic.levels.low.model_config]
-transport = "openai"
-model = "hermes-3"
-base_url = "http://ollama:11434/v1"
-
-[dialectic.levels.medium]
-MAX_TOOL_ITERATIONS = 2
-[dialectic.levels.medium.model_config]
-transport = "openai"
-model = "hermes-3"
-base_url = "http://ollama:11434/v1"
-
-[dialectic.levels.high]
-MAX_TOOL_ITERATIONS = 4
-[dialectic.levels.high.model_config]
-transport = "openai"
-model = "hermes-3"
-base_url = "http://ollama:11434/v1"
-
-[dialectic.levels.max]
-MAX_TOOL_ITERATIONS = 10
-[dialectic.levels.max.model_config]
-transport = "openai"
-model = "hermes-3"
-base_url = "http://ollama:11434/v1"
-
-# --- Summary ---
-[summary]
-ENABLED = true
-MESSAGES_PER_SHORT_SUMMARY = 20
-MESSAGES_PER_LONG_SUMMARY = 60
-
-[summary.model_config]
-transport = "openai"
-model = "hermes-3"
-base_url = "http://ollama:11434/v1"
-
-# --- Dream ---
-[dream]
-ENABLED = false
-
-# --- Peer Card ---
-[peer_card]
-ENABLED = true
-
-# --- Vector Store ---
-[vector_store]
-TYPE = "pgvector"
-DIMENSIONS = 768

ai/honcho/init-db.sql

@@ -1 +0,0 @@
-CREATE EXTENSION IF NOT EXISTS vector;