feat: add git-consistent memory gateway architecture

2026-03-07 22:33:41 +08:00
commit d4cd81f498
40 changed files with 2114 additions and 0 deletions
--- a/scripts/bootstrap.sh
+++ b/scripts/bootstrap.sh
@@ -0,0 +1,60 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
+cd "${ROOT_DIR}"
+
+mkdir -p data/git-mirror data/workspaces data/qmd-cache data/qmd-config
+
+REMOTE_BARE="${ROOT_DIR}/data/remote-memory.git"
+if [[ ! -d "${REMOTE_BARE}" ]]; then
+  echo "[bootstrap] initializing demo remote repo at ${REMOTE_BARE}"
+  git init --bare "${REMOTE_BARE}"
+
+  tmp_dir="$(mktemp -d)"
+  trap 'rm -rf "${tmp_dir}"' EXIT
+
+  git clone "${REMOTE_BARE}" "${tmp_dir}/seed"
+  git -C "${tmp_dir}/seed" config user.name "Memory Gateway Bot"
+  git -C "${tmp_dir}/seed" config user.email "memory-gateway@example.local"
+
+  mkdir -p "${tmp_dir}/seed/docs"
+  cat > "${tmp_dir}/seed/docs/main.md" <<'EOF'
+# Main Memory
+
+router firmware build recovery strategy in main branch.
+EOF
+
+  git -C "${tmp_dir}/seed" add docs/main.md
+  git -C "${tmp_dir}/seed" commit -m "init main memory"
+  git -C "${tmp_dir}/seed" branch -M main
+  git -C "${tmp_dir}/seed" push -u origin main
+
+  git -C "${tmp_dir}/seed" checkout -b memory/2026-03
+  cat > "${tmp_dir}/seed/docs/monthly.md" <<'EOF'
+# Monthly Memory 2026-03
+
+monthly branch note for March 2026.
+EOF
+  git -C "${tmp_dir}/seed" add docs/monthly.md
+  git -C "${tmp_dir}/seed" commit -m "add monthly memory"
+  git -C "${tmp_dir}/seed" push -u origin memory/2026-03
+
+  git -C "${tmp_dir}/seed" checkout -b task/TASK-001 main
+  cat > "${tmp_dir}/seed/docs/task-TASK-001.md" <<'EOF'
+# Task Memory TASK-001
+
+task specific recovery checkpoint.
+EOF
+  git -C "${tmp_dir}/seed" add docs/task-TASK-001.md
+  git -C "${tmp_dir}/seed" commit -m "add task memory"
+  git -C "${tmp_dir}/seed" push -u origin task/TASK-001
+fi
+
+if [[ ! -f .env ]]; then
+  cp .env.example .env
+  sed -i "s|^GIT_REMOTE_URL=.*$|GIT_REMOTE_URL=/data/remote-memory.git|" .env
+  echo "[bootstrap] generated .env with local demo remote: /data/remote-memory.git"
+fi
+
+echo "[bootstrap] done"
--- a/scripts/entrypoint.sh
+++ b/scripts/entrypoint.sh
@@ -0,0 +1,4 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+exec /app/scripts/qmd-entrypoint.sh "$@"
--- a/scripts/qmd-entrypoint.sh
+++ b/scripts/qmd-entrypoint.sh
@@ -0,0 +1,65 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+CACHE_HOME="${XDG_CACHE_HOME:-/var/lib/qmd/cache}"
+CONFIG_HOME="${XDG_CONFIG_HOME:-/var/lib/qmd/config}"
+PORT="${QMD_HTTP_PORT:-8181}"
+INTERNAL_PORT="${QMD_INTERNAL_PORT:-8182}"
+QMD_LLM_FILE="${QMD_LLM_FILE:-/usr/local/lib/node_modules/@tobilu/qmd/dist/llm.js}"
+
+mkdir -p "${CACHE_HOME}/qmd" "${CONFIG_HOME}/qmd"
+
+apply_model_overrides() {
+  if [[ -z "${QMD_EMBED_MODEL_URI:-}" && -z "${QMD_RERANK_MODEL_URI:-}" && -z "${QMD_GENERATE_MODEL_URI:-}" ]]; then
+    return 0
+  fi
+
+  if [[ ! -f "${QMD_LLM_FILE}" ]]; then
+    echo "WARN: llm.js not found at ${QMD_LLM_FILE}, skip model overrides." >&2
+    return 0
+  fi
+
+  QMD_LLM_FILE="${QMD_LLM_FILE}" node - <<'NODE'
+const fs = require("fs");
+const llmFile = process.env.QMD_LLM_FILE;
+let src = fs.readFileSync(llmFile, "utf8");
+
+const replacements = [
+  ["DEFAULT_EMBED_MODEL", process.env.QMD_EMBED_MODEL_URI],
+  ["DEFAULT_RERANK_MODEL", process.env.QMD_RERANK_MODEL_URI],
+  ["DEFAULT_GENERATE_MODEL", process.env.QMD_GENERATE_MODEL_URI],
+];
+
+for (const [name, value] of replacements) {
+  if (!value) continue;
+  const pattern = new RegExp(`const ${name} = \\"[^\\"]+\\";`);
+  if (!pattern.test(src)) {
+    console.error(`WARN: failed to find ${name} in ${llmFile}`);
+    continue;
+  }
+  src = src.replace(pattern, `const ${name} = ${JSON.stringify(value)};`);
+}
+
+fs.writeFileSync(llmFile, src, "utf8");
+NODE
+}
+
+apply_model_overrides
+
+qmd mcp --http --port "${INTERNAL_PORT}" &
+qmd_pid=$!
+
+socat "TCP-LISTEN:${PORT},fork,reuseaddr,bind=0.0.0.0" "TCP6:[::1]:${INTERNAL_PORT}" &
+proxy_pid=$!
+
+cleanup() {
+  kill "${proxy_pid}" "${qmd_pid}" 2>/dev/null || true
+  wait "${proxy_pid}" "${qmd_pid}" 2>/dev/null || true
+}
+
+trap cleanup SIGINT SIGTERM
+
+wait -n "${qmd_pid}" "${proxy_pid}"
+status=$?
+cleanup
+exit "${status}"
--- a/scripts/smoke-test.sh
+++ b/scripts/smoke-test.sh
@@ -0,0 +1,29 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
+cd "${ROOT_DIR}"
+
+log() {
+  printf '[smoke] %s\n' "$*"
+}
+
+log "Bootstrapping demo repo"
+./scripts/bootstrap.sh
+
+log "Starting services"
+docker compose up -d --build
+
+log "Checking health"
+curl -fsS http://127.0.0.1:8181/health >/dev/null
+curl -fsS http://127.0.0.1:8787/health >/dev/null
+
+log "Running query through memory-gateway"
+query_output="$(curl -fsS -X POST http://127.0.0.1:8787/query \
+  -H 'content-type: application/json' \
+  -d '{"branch":"main","query_type":"search","query":"router","require_latest":true,"debug":true}')"
+printf '%s\n' "${query_output}"
+printf '%s\n' "${query_output}" | grep -F 'commit_hash' >/dev/null
+printf '%s\n' "${query_output}" | grep -F 'branch' >/dev/null
+
+log "Smoke test passed"
--- a/scripts/warmup.sh
+++ b/scripts/warmup.sh
@@ -0,0 +1,27 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
+cd "${ROOT_DIR}"
+
+BASE_URL="${GATEWAY_BASE_URL:-http://127.0.0.1:8787}"
+
+wait_health() {
+  local attempts=60
+  for _ in $(seq 1 "${attempts}"); do
+    if curl -fsS "${BASE_URL}/health" >/dev/null 2>&1; then
+      return 0
+    fi
+    sleep 2
+  done
+  echo "ERROR: gateway did not become healthy at ${BASE_URL}/health" >&2
+  return 1
+}
+
+wait_health
+
+curl -fsS -X POST "${BASE_URL}/query" \
+  -H 'content-type: application/json' \
+  -d '{"branch":"main","query_type":"search","query":"router","require_latest":true,"debug":true}' >/dev/null
+
+echo "Warmup completed."