feat: add llama-cpp-hermes service with ROCm 6.1 + gfx906 support

- Add custom llama.cpp Dockerfile with ROCm 6.1 + gfx906 (MI50) build
- Add llama-cpp-hermes service serving Hermes 4.3 on dual MI50 GPUs
- Strip GPU devices/ROCm env from ollama service (CPU-only for embeddings)

Hermes 4.3 runs at ~19 t/s on dual MI50s with 160K context.

.gitea/workflows/build-hermes.yml

@@ -0,0 +1,31 @@
+name: Build Hermes agent
+on:
+  pull_request:
+    branches: [ master ]
+    paths:
+      - 'ai/hermes/**'
+      - 'ai/compose.yml'
+  push:
+    branches: [ master ]
+    paths:
+      - 'ai/hermes/**'
+      - 'ai/compose.yml'
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        run: |
+          git clone -b "${{ github.head_ref || github.ref_name }}" \
+            https://gitea:${{ secrets.GITHUB_TOKEN }}@code.lazyworkhorse.net/gortium/compose.git .
+          git log --oneline -3
+
+      - name: Build hermes image
+        run: |
+          cd ai
+          docker compose build hermes 2>&1
+
+      - name: Verify image
+        run: |
+          docker run --rm ai-hermes /opt/hermes/.venv/bin/python --version 2>&1

.gitea/workflows/build-ollama.yml

@@ -0,0 +1,31 @@
+name: Build ollama (gfx906)
+on:
+  pull_request:
+    branches: [ master ]
+    paths:
+      - 'ai/ollama/**'
+      - 'ai/compose.yml'
+  push:
+    branches: [ master ]
+    paths:
+      - 'ai/ollama/**'
+      - 'ai/compose.yml'
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        run: |
+          git clone -b "${{ github.head_ref || github.ref_name }}" \
+            https://gitea:${{ secrets.GITHUB_TOKEN }}@code.lazyworkhorse.net/gortium/compose.git .
+          git log --oneline -3
+
+      - name: Build ollama image
+        run: |
+          cd ai
+          docker compose build ollama --no-cache 2>&1
+
+      - name: Verify version
+        run: |
+          docker run --rm ollama/ollama:rocm-gfx906 ollama --version 2>&1

ai/compose.yml

@@ -31,6 +31,9 @@ services:
       ssh:
         - default
     container_name: hermes
+    entrypoint: ["/bin/bash", "-c",
+      "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
     command: gateway run
@@ -51,6 +54,10 @@ services:
       - TZ=America/Montreal
     volumes:
       - /mnt/HoardingCow_docker_data/Hermes/data:/opt/data
+      # Syncthing-shared org files — read-only view of user's agenda
+      - /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
     devices:
       - /dev/kfd:/dev/kfd
       - /dev/dri:/dev/dri
@@ -60,6 +67,35 @@ services:
     networks:
       - ai_backend
 
+  syncthing:
+    image: syncthing/syncthing:latest
+    container_name: syncthing
+    hostname: syncthing
+    restart: always
+    ports:
+      - "8384:8384"
+      - "22000:22000"
+      - "21027:21027/udp"
+    environment:
+      - TZ=America/Montreal
+    volumes:
+      - /mnt/HoardingCow_docker_data/Syncthing/config:/var/syncthing/config
+      - /mnt/HoardingCow_docker_data/Syncthing/telos-ro:/telos-ro
+      - /mnt/HoardingCow_docker_data/Syncthing/telos-rw:/telos-rw
+    networks:
+      - ai_backend
+      - ai_net
+    labels:
+      - "traefik.enable=true"
+      - "traefik.http.routers.syncthing-http.rule=Host(`syncthing.lazyworkhorse.net`)"
+      - "traefik.http.routers.syncthing-http.entrypoints=web"
+      - "traefik.http.routers.syncthing-http.middlewares=redirect-to-https"
+      - "traefik.http.routers.syncthing-https.rule=Host(`syncthing.lazyworkhorse.net`)"
+      - "traefik.http.routers.syncthing-https.entrypoints=websecure"
+      - "traefik.http.routers.syncthing-https.tls=true"
+      - "traefik.http.routers.syncthing-https.tls.certresolver=njalla"
+      - "traefik.http.services.syncthing.loadbalancer.server.port=8384"
+
   ollama:
     build:
       context: ./ollama
@@ -76,22 +112,7 @@ services:
       - /mnt/HoardingCow_docker_data/Ollama/ollama:/root/.ollama
     environment:
       - OLLAMA_VULKAN=0
-      - HSA_OVERRIDE_GFX_VERSION=9.0.6
-      - HCC_AMDGPU_TARGET=gfx906
-      - HIP_VISIBLE_DEVICES=0,1
-      - ROCR_VISIBLE_DEVICES=0,1
-      - HSA_ENABLE_SDMA=0 
       - OLLAMA_HOST=0.0.0.0
-      - OLLAMA_DEBUG=1
-      - OLLAMA_FLASH_ATTENTION=1
-      - OLLAMA_NUM_PARALLEL=2
-    devices:
-      # Map the render nodes and KFD for ROCm to work inside the container
-      - /dev/kfd:/dev/kfd
-      - /dev/dri:/dev/dri
-    group_add:
-      - "303"
-      - "26"
 
 networks:
   ai_net:
@@ -101,47 +122,40 @@ networks:
     driver: bridge
     name: ai_backend
     
-  # llama_cpp_devstral:
-  #   image: ghcr.io/ggml-org/llama.cpp:server-rocm
-  #   container_name: llama_cpp_devstral
-  #   restart: unless-stopped
-  #   networks:
-  #     - ai_backend
-  #   ports:
-  #     - "8300:8080"
-  #   ipc: host
-  #   devices:
-  #     - "/dev/kfd:/dev/kfd"
-  #     - "/dev/dri:/dev/dri"
-  #   group_add:
-  #     - "303" # video
-  #     - "26"  # render
-  #   environment:
-  #     HSA_OVERRIDE_GFX_VERSION: 9.0.6
-  #     HIP_VISIBLE_DEVICES: 0,1
-  #     LLAMA_CACHE: /models
-  #   volumes:
-  #     - /mnt/HoardingCow_docker_data/Llama_cpp/models:/models
-  #     - /mnt/HoardingCow_docker_data/Llama_cpp/devstral-agent.jinja:/template.jinja
-  #   command: >
-  #     -hf unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF:Devstral-Small-2-24B-Instruct-2512-Q8_0.gguf
-  #     -a devstral-2-small-llama_cpp
-  #     --chat-template-file /template.jinja
-  #     --host 0.0.0.0
-  #     --port 8080
-  #     --n-gpu-layers 99
-  #     --ctx-size 163840
-  #     --batch-size 4096
-  #     --ubatch-size 4096
-  #     --cache-type-k f16
-  #     --cache-type-v f16
-  #     --cache-reuse 256
-  #     --flash-attn on
-  #     --context-shift
-  #     --split-mode layer
-  #     --no-mmap
-  #     --n-predict -1
-  #     --parallel 2
+  llama-cpp-hermes:
+    image: llama-cpp:rocm-gfx906
+    container_name: llama-cpp-hermes
+    restart: unless-stopped
+    networks:
+      - ai_backend
+    ports:
+      - "127.0.0.1:8300:8080"
+    ipc: host
+    devices:
+      - /dev/kfd:/dev/kfd
+      - /dev/dri:/dev/dri
+    group_add:
+      - "303"
+      - "26"
+    environment:
+      - HSA_OVERRIDE_GFX_VERSION=9.0.6
+      - HSA_ENABLE_SDMA=0
+      - HIP_VISIBLE_DEVICES=0,1
+      - LLAMA_CACHE=/models
+    volumes:
+      - /mnt/HoardingCow_docker_data/Llama_cpp/models:/models
+      - /mnt/HoardingCow_docker_data/Ollama/ollama/models/blobs/sha256-17823599694fa3503ef54bf748d5078c6ce881f4d01616cafa255dc05d215a08:/model.gguf:ro
+    command: >
+      -m /model.gguf
+      --host 0.0.0.0
+      --port 8080
+      --gpu-layers 99
+      --ctx-size 163840
+      -ctk f16 -ctv f16
+      --flash-attn on
+      --split-mode layer
+      --no-mmap
+      --n-predict -1
 
   # vllm:
   #   image: nalanzeyu/vllm-gfx906:v0.9.0-rocm6.3

ai/hermes/Dockerfile

@@ -61,6 +61,24 @@ urllib.request.urlretrieve(url, base + '/en_US-ryan-high.onnx')
 urllib.request.urlretrieve(url + '.json', base + '/en_US-ryan-high.onnx.json')
 PYEOF
 
+# ---------- Install Himalaya email CLI ----------
+RUN /opt/hermes/.venv/bin/python3 /dev/stdin << 'PYEOF'
+import urllib.request, tarfile, os, shutil
+url = 'https://github.com/pimalaya/himalaya/releases/download/v1.2.0/himalaya.x86_64-linux.tgz'
+tgz = '/tmp/himalaya.tgz'
+urllib.request.urlretrieve(url, tgz)
+with tarfile.open(tgz) as t:
+    t.extractall('/tmp')
+shutil.move('/tmp/himalaya', '/usr/local/bin/himalaya')
+os.chmod('/usr/local/bin/himalaya', 0o755)
+os.remove(tgz)
+print('himalaya v1.2.0 installed')
+PYEOF
+
+# ---------- Install himalaya-ro wrapper ----------
+COPY --chmod=0755 himalaya-ro.sh /usr/local/bin/himalaya-ro
+
+
 # ---------- Runtime ----------
 USER hermes
 ENV HERMES_HOME=/opt/data
@@ -68,7 +86,8 @@ ENV PATH="/opt/data/.local/bin:${PATH}"
 # Point browser tool to Playwright's Chromium (already in base image)
 ENV CHROME_EXECUTABLE=/opt/hermes/.playwright/chromium/chrome-linux/chrome
 
-VOLUME [ "/opt/data" ]
+# Ensure tools directory and toolsets.py are writable by the hermes runtime user
+# so custom tools can be injected from the persistent volume at startup.
+RUN chown -R hermes:hermes /opt/hermes/tools /opt/hermes/toolsets.py
 
-COPY --chmod=0755 fix-permissions.sh /opt/hermes/fix-permissions.sh
-ENTRYPOINT [ "/usr/bin/tini", "-g", "--", "/opt/hermes/fix-permissions.sh" ]
+VOLUME [ "/opt/data" ]

ai/hermes/fix-permissions.sh

@@ -1,38 +0,0 @@
-#!/bin/bash
-# Startup permission fix + TTS patch.
-# Runs as root before the entrypoint drops to the hermes user.
-set -e
-
-HERMES_HOME="${HERMES_HOME:-/opt/data}"
-
-# Fix ownership on critical writable directories
-chown -R hermes:hermes \
-  "$HERMES_HOME/sessions" \
-  "$HERMES_HOME/checkpoints" \
-  "$HERMES_HOME/skills" \
-  "$HERMES_HOME/memories" \
-  "$HERMES_HOME/workspace" \
-  "$HERMES_HOME/pastes" \
-  "$HERMES_HOME/logs" \
-  "$HERMES_HOME/cron" \
-  "$HERMES_HOME/plans" \
-  "$HERMES_HOME/hooks" \
-  "$HERMES_HOME/cache" \
-  2>/dev/null || true
-
-# Fix data volume root ownership
-if [ "$(stat -c %u "$HERMES_HOME" 2>/dev/null)" != "$(id -u hermes)" ]; then
-  chown hermes:hermes "$HERMES_HOME" 2>/dev/null || true
-fi
-
-# ---------- Patch tts_tool.py: replace Edge TTS with Piper ----------
-# Fallback runtime patch in case the volume's site-packages differ from the image.
-# Idempotent: if already patched, the script does nothing.
-PATCH_SCRIPT="/opt/hermes/patch_tts_tool.py"
-if [ -f "$PATCH_SCRIPT" ]; then
-  echo "Applying TTS patch (Piper only, no Edge fallback)..."
-  /opt/hermes/.venv/bin/python3 "$PATCH_SCRIPT" 2>&1 || true
-fi
-
-# Chain to the official Hermes entrypoint
-exec /opt/hermes/docker/entrypoint.sh "$@"

ai/hermes/himalaya-ro.sh

@@ -0,0 +1,73 @@
+#!/usr/bin/env bash
+# ─────────────────────────────────────────────────────────────
+# himalaya-ro — Read-only wrapper for himalaya
+#
+# Blocks destructive commands and logs audit trail.
+# Pass-through for read-only commands (list, read, search).
+#
+# Usage:  himalaya-ro [options] <command> [args...]
+#
+# Install: place in PATH before the real himalaya, or use
+#          `ln -sf himalaya-ro /usr/local/bin/himalaya`
+# ─────────────────────────────────────────────────────────────
+set -o pipefail
+
+# ── Configuration ───────────────────────────────────────────
+HIMALAYA_BIN="${HIMALAYA_BIN:-/usr/local/bin/himalaya}"
+AUDIT_LOG="${HIMALAYA_AUDIT_LOG:-/var/log/himalaya-audit.log}"
+
+# ── Destructive commands we block ──────────────────────────
+BLOCKED_CMDS=(
+  "message move"
+  "message delete"
+  "message copy"
+  "flag add"
+  "flag remove"
+  "folder create"
+  "folder delete"
+  "folder rename"
+  "template send"
+  "account configure"
+  "account delete"
+)
+
+# ── Determine the subcommand being invoked ─────────────────
+# Strip leading options (--account, --output, etc.) to find the verb
+ARGS=()
+SKIP_NEXT=false
+for arg in "$@"; do
+  if $SKIP_NEXT; then
+    SKIP_NEXT=false
+    continue
+  fi
+  if [[ "$arg" == --* ]]; then
+    case "$arg" in
+      --account|--output|--page|--page-size|--folder|--color|--format)
+        SKIP_NEXT=true ;;
+    esac
+    continue
+  fi
+  ARGS+=("$arg")
+done
+
+# Build subcommand string and check against blocklist
+CMD_STR=""
+for ((i=0; i<${#ARGS[@]}; i++)); do
+  if [ -z "$CMD_STR" ]; then
+    CMD_STR="${ARGS[$i]}"
+  else
+    CMD_STR="$CMD_STR ${ARGS[$i]}"
+  fi
+  for blocked in "${BLOCKED_CMDS[@]}"; do
+    if [[ "$CMD_STR" == "$blocked" ]]; then
+      TS=$(date '+%Y-%m-%d %H:%M:%S')
+      echo "[AUDIT] $TS BLOCKED: himalaya $*" >> "$AUDIT_LOG"
+      echo "ERROR: Command 'himalaya $CMD_STR ...' is blocked by read-only policy." >&2
+      echo "       Audit log: $AUDIT_LOG" >&2
+      exit 100
+    fi
+  done
+done
+
+# ── Allow pass-through ─────────────────────────────────────
+exec "$HIMALAYA_BIN" "$@"

ai/llama-cpp/Dockerfile

@@ -0,0 +1,30 @@
+# llama-cpp-rocm6/Dockerfile
+# Custom llama.cpp server with ROCm 6.1 + gfx906 (MI50) support.
+# Build: docker build -t llama-cpp:rocm-gfx906 .
+
+FROM rocm/dev-ubuntu-22.04:6.1.2-complete AS builder
+RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y curl git build-essential pkg-config cmake make && rm -rf /var/lib/apt/lists/*
+ARG LLAMACPP_VERSION=b9596
+RUN git clone --depth 1 --branch ${LLAMACPP_VERSION} https://github.com/ggml-org/llama.cpp.git /build
+WORKDIR /build
+ENV HIP_PATH=/opt/rocm ROCM_PATH=/opt/rocm PATH=/opt/rocm/bin:/opt/rocm/llvm/bin:${PATH} CMAKE_PREFIX_PATH=/opt/rocm
+RUN mkdir build && cd build && \
+    cmake .. -DGGML_HIP=ON -DCMAKE_BUILD_TYPE=Release \
+      -DAMDGPU_TARGETS="gfx906:xnack-" \
+      -DCMAKE_POSITION_INDEPENDENT_CODE=ON \
+      -DGGML_CUDA=OFF -DGGML_VULKAN=OFF -DGGML_METAL=OFF \
+      -DBUILD_SHARED_LIBS=OFF && \
+    cmake --build . --target llama-server -- -j $(nproc)
+
+FROM ubuntu:24.04
+RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y \
+    ca-certificates curl libstdc++6 libgomp1 libopenblas0 \
+    libnuma1 libelf1 libdrm2 libdrm-amdgpu1 \
+    && rm -rf /var/lib/apt/lists/*
+COPY --from=builder /opt/rocm/lib/ /opt/rocm/lib/
+COPY --from=builder /opt/rocm/share/ /opt/rocm/share/
+COPY --from=builder /build/build/bin/llama-server /usr/local/bin/llama-server
+RUN echo /opt/rocm/lib > /etc/ld.so.conf.d/rocm.conf && ldconfig
+ENV HSA_OVERRIDE_GFX_VERSION=9.0.6 HCC_AMDGPU_TARGET=gfx906 HSA_ENABLE_SDMA=0
+EXPOSE 8080
+ENTRYPOINT ["/usr/local/bin/llama-server"]

versioncontrol/compose.yml

@@ -7,12 +7,11 @@ services:
       - USER_UID=1000
       - USER_GID=1000
       - GITEA__server__ROOT_URL=https://code.lazyworkhorse.net
+      - GITEA__actions__ENABLED=true
       - SSH_PORT=2222
       - SSH_LISTEN_PORT=2222
       # Enable Gitea Actions (act_runner required on host)
       - GITEA__actions__ENABLED=true
-      # Don't fetch actions from GitHub (offline mode + local only)
-      - GITEA__actions__DEFAULT_ACTIONS_URL=off
     volumes:
       - /mnt/HoardingCow_docker_data/Gitea:/data
     networks:
@@ -44,6 +43,22 @@ services:
       # Internal Routing
       - "traefik.http.services.gitea.loadbalancer.server.port=3000"
 
+  act_runner:
+    image: gitea/act_runner:latest
+    container_name: act_runner
+    environment:
+      - GITEA_INSTANCE_URL=https://code.lazyworkhorse.net
+      - GITEA_RUNNER_REGISTRATION_TOKEN=${GITEA_RUNNER_TOKEN}
+      - GITEA_RUNNER_NAME=ai-host-runner
+      - GITEA_RUNNER_LABELS=ubuntu-latest:docker://catthehacker/ubuntu:full-22.04,nixos-builder:docker://nixos/nix
+    volumes:
+      - /var/run/docker.sock:/var/run/docker.sock
+    networks:
+      - vc_net
+    restart: always
+    depends_on:
+      - gitea
+
 networks:
   vc_net:
     external: true

vpn/Dockerfile

@@ -0,0 +1,9 @@
+# Custom wg-easy with iptables-nft (nftables-backed iptables)
+# Fixes crash-loop when host kernel lacks legacy iptable_nat module.
+FROM ghcr.io/wg-easy/wg-easy:latest
+
+# The upstream image registers only iptables-legacy with update-alternatives.
+# iptables-nft binary exists but isn't registered as an alternative key.
+# Override the alternatives-managed symlinks directly.
+RUN ln -sf /usr/sbin/iptables-nft /usr/sbin/iptables && \
+    ln -sf /usr/sbin/ip6tables-nft /usr/sbin/ip6tables

vpn/compose.yml

@@ -2,7 +2,10 @@ version: "3.8"
 
 services:
   wireguard:
-    image: weejewel/wg-easy:latest
+    build:
+      context: .
+      dockerfile: Dockerfile
+    image: wg-easy-iptables-nft:latest
     container_name: wireguard
     cap_add:
       - NET_ADMIN