shell: kill process group on timeout/cancel (no orphans) + shell_bg job registry + kill_job/list_jobs

Root cause of the shell->shell_bg thrash: run() backgrounded POSIX commands
with a bare & and never captured the pid, so on timeout it returned an error
while the children kept running. Now: setsid (own process group), capture the
leader pid, SIGTERM+SIGKILL the whole group on timeout/cancel, and a message
telling the model it was killed + to use shell_bg. Plus shell_bg registers jobs
to /tmp/lmcp-bg-jobs.tsv and new kill_job/list_jobs let a runaway job be reaped
without a reboot.

.gitignore

@@ -0,0 +1,13 @@
+# Generated by windows/sync.sh — see windows/README.md
+windows/pkg/lmcp.lua
+windows/pkg/server.lua
+windows/pkg/json.lua
+
+# Bundled Lua + LuaSocket runtime for the Windows MSI; downloaded
+# separately, not in git.
+windows/pkg/lua/
+
+# Editor / OS noise
+*.swp
+*.swo
+.DS_Store

example_server.lua

@@ -11,6 +11,10 @@ local server = lmcp.new("example-tools", {
     port = tonumber(arg[1]) or 8080,
 })
 
+-- The optional 5th `opts` arg to server:tool carries MCP annotations.
+-- Omit it and clients assume the worst (destructive, openWorld) — fine
+-- for prototypes; declare annotations once you know each tool's stance.
+
 server:tool("shell", "Execute a shell command", {
     type = "object",
     properties = {
@@ -24,7 +28,15 @@ server:tool("shell", "Execute a shell command", {
     local result = handle:read('*a')
     handle:close()
     return result ~= '' and result or '(no output)'
-end)
+end, {
+    annotations = {
+        title           = "Run shell",
+        readOnlyHint    = false,
+        destructiveHint = true,
+        idempotentHint  = false,
+        openWorldHint   = true,
+    },
+})
 
 server:tool("read_file", "Read a file", {
     type = "object",
@@ -38,7 +50,15 @@ server:tool("read_file", "Read a file", {
     local content = f:read('*a')
     f:close()
     return content
-end)
+end, {
+    annotations = {
+        title           = "Read file",
+        readOnlyHint    = true,
+        destructiveHint = false,
+        idempotentHint  = true,
+        openWorldHint   = false,
+    },
+})
 
 server:tool("write_file", "Write content to a file", {
     type = "object",
@@ -53,7 +73,15 @@ server:tool("write_file", "Write content to a file", {
     f:write(args.content)
     f:close()
     return string.format("Written %d bytes to %s", #args.content, args.path)
-end)
+end, {
+    annotations = {
+        title           = "Write file",
+        readOnlyHint    = false,
+        destructiveHint = true,
+        idempotentHint  = true,
+        openWorldHint   = false,
+    },
+})
 
 server:tool("list_dir", "List directory contents", {
     type = "object",
@@ -67,7 +95,85 @@ server:tool("list_dir", "List directory contents", {
     local result = handle:read('*a')
     handle:close()
     return result
+end, {
+    annotations = {
+        title           = "List directory",
+        readOnlyHint    = true,
+        destructiveHint = false,
+        idempotentHint  = true,
+        openWorldHint   = false,
+    },
+})
+
+-- ---- Resources (MCP primitive — see issue #5) ----
+-- Tools-only servers force the client to spend a tools/call round-trip
+-- for every read. Resources let the client list and read by URI, with a
+-- stable identity it can cache and reference in prompts.
+
+server:resource("text://greeting", {
+    name = "Greeting",
+    mimeType = "text/plain",
+}, function() return "Hello from lmcp!" end)
+
+-- Tiny binary resource: 8-byte PNG signature, demonstrates blob handling.
+server:resource("data://lmcp.png", {
+    name = "PNG signature",
+    mimeType = "image/png",
+}, function()
+    return { blob_bytes = "\x89PNG\r\n\x1a\n", mimeType = "image/png" }
 end)
 
-io.stderr:write("Starting lmcp example server...\n")
-server:run()
+-- Template: any local file. `args.path` is captured greedily (no leading
+-- slash because the template literal already includes ///).
+server:resource_template("file:///{path}", {
+    name = "Local file",
+    mimeType = "text/plain",
+}, function(args)
+    local f = io.open("/" .. args.path, "r")
+    if not f then error("file not found: /" .. args.path) end
+    local content = f:read("*a"); f:close()
+    return content
+end)
+
+-- ---- Prompts (MCP primitive — see issue #6) ----
+-- Parameterised prompt templates the client surfaces as a menu
+-- (slash-commands, snippets). Handler returns either a plain string (one
+-- user-role text message) or a full { description?, messages = {...} }
+-- shape for finer control.
+
+server:prompt("release_note", {
+    description = "Draft a release note for a given version",
+    arguments = {
+        { name = "version", description = "Tag, e.g. v0.7.1", required = true },
+        { name = "since",   description = "Previous tag",      required = false },
+    },
+}, function(args)
+    return "Write concise release notes for version " .. (args.version or "?")
+        .. " since " .. (args.since or "the previous tag")
+        .. ". Group by category (features / fixes / docs)."
+end)
+
+-- Completion for the release_note prompt's `version` argument. Returned
+-- list is filtered against `value` (prefix match) by the server's spec
+-- contract is "candidates"; clients may further filter.
+server:complete("ref/prompt", "release_note", "version", function(value, ctx)
+    local all = { "v0.5.0", "v0.5.1", "v0.5.2", "v0.5.3", "v0.5.4",
+                  "v0.6.0", "v0.7.0", "v0.7.1", "v1.0.0-rc1" }
+    if value == "" then return all end
+    local out = {}
+    for _, v in ipairs(all) do
+        if v:sub(1, #value) == value then out[#out + 1] = v end
+    end
+    return out
+end)
+
+local transport = os.getenv("LMCP_TRANSPORT") or "http"
+if transport == "stdio" then
+    if os.getenv("LMCP_PORT") then
+        io.stderr:write("lmcp: LMCP_PORT ignored in stdio mode\n")
+    end
+    server:run_stdio()
+else
+    io.stderr:write("Starting lmcp example server...\n")
+    server:run()
+end

examples/lmcp.service

@@ -0,0 +1,22 @@
+[Unit]
+Description=lmcp MCP Server
+After=network.target
+
+[Service]
+Type=simple
+User=root
+# Arch ships the Lua 5.4 binary as /usr/bin/lua; Debian ships /usr/bin/lua5.4.
+# Override ExecStart if your distro differs.
+ExecStart=/usr/bin/lua /usr/share/lua/5.4/server.lua
+# Distinct name per host: foo-tools appears in /mcp listings and logs.
+Environment=LMCP_NAME=CHANGEME-tools
+Environment=LMCP_PORT=8080
+# Bearer token. Generate with: openssl rand -hex 24
+# For untrusted networks, bind to LAN-only via firewall; the server itself
+# listens on 0.0.0.0 by default.
+Environment=LMCP_TOKEN=CHANGEME
+Restart=on-failure
+RestartSec=5
+
+[Install]
+WantedBy=multi-user.target

helpers/stash_recall.py

@@ -0,0 +1,29 @@
+import urllib.request, json, threading, queue, sys
+BASE="http://192.168.88.184:8080"
+q=queue.Queue()
+def sse():
+    try:
+        r=urllib.request.urlopen(BASE+"/sse", timeout=30)
+        for raw in r:
+            s=raw.decode(errors="replace").strip()
+            if s.startswith("data:"): q.put(s[5:].strip())
+    except Exception as e: q.put("ERR:"+str(e))
+threading.Thread(target=sse,daemon=True).start()
+try:
+    ep=q.get(timeout=10)
+    if ep.startswith("ERR:"): print("stash unreachable:",ep); sys.exit(1)
+    purl=BASE+ep if ep.startswith("/") else ep
+    def post(o):
+        urllib.request.urlopen(urllib.request.Request(purl,data=json.dumps(o).encode(),headers={"Content-Type":"application/json"}),timeout=15).read()
+    post({"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"apropos","version":"1"}}})
+    q.get(timeout=10)
+    post({"jsonrpc":"2.0","method":"notifications/initialized"})
+    query=sys.argv[1] if len(sys.argv)>1 else ""
+    limit=int(sys.argv[2]) if len(sys.argv)>2 else 3
+    post({"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"recall","arguments":{"query":query,"limit":limit}}})
+    d=json.loads(q.get(timeout=25))
+    txt=d.get("result",{}).get("content",[{}])[0].get("text","[]")
+    facts=json.loads(txt)
+    if not facts: print("(no memory found for: %s)"%query); sys.exit(0)
+    for f in facts: print("- %s (score %.2f)" % (f.get("content","").strip(), f.get("score",0)))
+except Exception as e: print("recall error:",e); sys.exit(1)

hub.lua

@@ -25,9 +25,26 @@ local socket = require('socket')
 -- ---- Backend registry ---------------------------------------------------
 
 local CONF_PATH = os.getenv("LMCP_HUB_BACKENDS") or "/opt/herding/etc/hub-backends.conf"
-local PROBE_TTL = tonumber(os.getenv("LMCP_HUB_PROBE_TTL") or "30")
+local PROBE_TTL_UP = tonumber(os.getenv("LMCP_HUB_PROBE_TTL_UP") or "30")
+local PROBE_TTL_DOWN_MIN = tonumber(os.getenv("LMCP_HUB_PROBE_TTL_DOWN_MIN") or "60")
+local PROBE_TTL_DOWN_MAX = tonumber(os.getenv("LMCP_HUB_PROBE_TTL_DOWN_MAX") or "900")
+local PROBE_BUDGET = tonumber(os.getenv("LMCP_HUB_PROBE_BUDGET") or "3")
 local LMCP_TIMEOUT = tonumber(os.getenv("LMCP_HUB_LMCP_TIMEOUT") or "6")
 local SSH_TIMEOUT = tonumber(os.getenv("LMCP_HUB_SSH_TIMEOUT") or "10")
+local SSH_HARD_TIMEOUT = tonumber(os.getenv("LMCP_HUB_SSH_HARD_TIMEOUT") or "30")
+local LOG_REQUESTS = (os.getenv("LMCP_HUB_LOG") or "1") ~= "0"
+
+local function logreq(fmt, ...)
+    if not LOG_REQUESTS then return end
+    io.stderr:write(string.format("[hub %s] " .. fmt .. "\n",
+        os.date("%H:%M:%S"), ...))
+    io.stderr:flush()
+end
+
+local function monotonic()
+    local ok, s = pcall(socket.gettime)
+    return ok and s or os.time()
+end
 
 local backends = {}  -- name -> { name, ssh_host, lmcp_url, token }
 local status = {}    -- name -> { up=bool, via="lmcp"|"ssh"|nil, checked=t, err=... }
@@ -160,12 +177,17 @@ local function ssh_run_script(ssh_host, script)
     local spath, perr = tmpwrite(script)
     if not spath then return nil, perr, -1 end
     local outpath = spath .. ".out"
+    -- Hard-cap wall time with GNU `timeout` — ssh's ConnectTimeout only
+    -- bounds TCP connect, not the session. Without this, a half-dead sshd
+    -- (auth stall, remote bash-s hang) locks the hub event loop indefinitely.
     local cmd = string.format(
-        "ssh -o ConnectTimeout=%d -o BatchMode=yes -o StrictHostKeyChecking=accept-new %s 'bash -s' < %s > %s 2>&1; echo $? > %s.rc",
-        SSH_TIMEOUT, shell_quote(ssh_host), shell_quote(spath),
+        "timeout --kill-after=2 %d ssh -o ConnectTimeout=%d -o ServerAliveInterval=5 -o ServerAliveCountMax=2 -o BatchMode=yes -o StrictHostKeyChecking=accept-new %s 'bash -s' < %s > %s 2>&1; echo $? > %s.rc",
+        SSH_HARD_TIMEOUT, SSH_TIMEOUT, shell_quote(ssh_host), shell_quote(spath),
         shell_quote(outpath), shell_quote(spath)
     )
+    local t0 = monotonic()
     os.execute(cmd)
+    local dt = monotonic() - t0
     local rc_f = io.open(spath .. ".rc", "r")
     local rc = -1
     if rc_f then
@@ -178,42 +200,123 @@ local function ssh_run_script(ssh_host, script)
     local output = out_f and out_f:read("*a") or ""
     if out_f then out_f:close() end
     os.remove(spath); os.remove(outpath); os.remove(spath .. ".rc")
+    -- timeout(1) exits 124 on wall-clock expiry, 137 on SIGKILL. Surface it.
+    if rc == 124 or rc == 137 then
+        logreq("ssh HARD-TIMEOUT host=%s after=%.2fs rc=%d", ssh_host, dt, rc)
+        return nil, string.format("ssh hard-timeout after %ds", SSH_HARD_TIMEOUT), rc
+    end
+    logreq("ssh host=%s elapsed=%.2fs rc=%d bytes=%d", ssh_host, dt, rc, #output)
     return output, nil, rc
 end
 
 -- ---- Health probe ------------------------------------------------------
+--
+-- Design notes:
+--   - Probe is lmcp-only. SSH is NOT checked here: it's expensive (3-6s per
+--     offline host) and the hub exists specifically to absorb lots of
+--     offline hosts. Hosts with lmcp down but ssh up show as DOWN in the
+--     health list, but actual remote_* calls still fall through to ssh
+--     fallback correctly.
+--   - Sticky DOWN cache with exponential backoff: a DOWN host is re-probed
+--     at intervals that grow from PROBE_TTL_DOWN_MIN (default 60s) up to
+--     PROBE_TTL_DOWN_MAX (default 900s). Prevents a sleeping fleet from
+--     burning probe budget every health check.
+--   - `remote_list_hosts` uses a parallel curl fan-out for all hosts at
+--     once so wall-clock is bounded by PROBE_BUDGET (default 3s), not the
+--     sum of per-host timeouts.
 
+local function cache_fresh(s, now)
+    if not s then return false end
+    local ttl
+    if s.up then
+        ttl = PROBE_TTL_UP
+    else
+        ttl = math.min(PROBE_TTL_DOWN_MAX, PROBE_TTL_DOWN_MIN * (2 ^ (s.down_streak or 0)))
+    end
+    return (now - s.checked) < ttl
+end
+
+local function apply_probe_result(name, up, err, via, tool_count)
+    local prev = status[name]
+    local new_s = {
+        checked = os.time(),
+        up = up,
+        via = up and (via or "lmcp") or nil,
+        err = err,
+        tool_count = tool_count,
+        down_streak = up and 0 or ((prev and prev.down_streak or 0) + 1),
+    }
+    status[name] = new_s
+    return new_s
+end
+
+-- Fallback single-host probe used only when a caller explicitly needs a
+-- status refresh outside the parallel path (currently only
+-- remote_list_hosts when a single host is looked up). lmcp-only.
 local function probe(name, force)
     local b = backends[name]
     if not b then return { up = false, err = "unknown host" } end
-    local now = os.time()
     local s = status[name]
-    if not force and s and (now - s.checked) < PROBE_TTL then
-        return s
+    if not force and cache_fresh(s, os.time()) then return s end
+    if not b.lmcp_url then
+        return apply_probe_result(name, false, "no lmcp url", nil, nil)
     end
-    local new_s = { checked = now, up = false, via = nil, err = nil }
-    if b.lmcp_url then
-        local result, err = jsonrpc_call(b.lmcp_url, b.token, "tools/list", {})
-        if result and result.tools then
-            new_s.up = true
-            new_s.via = "lmcp"
-            new_s.tool_count = #result.tools
-        else
-            new_s.err = "lmcp: " .. tostring(err)
+    local result, err = jsonrpc_call(b.lmcp_url, b.token, "tools/list", {})
+    if result and result.tools then
+        return apply_probe_result(name, true, nil, "lmcp", #result.tools)
+    end
+    return apply_probe_result(name, false, "lmcp: " .. tostring(err), nil, nil)
+end
+
+-- Parallel lmcp probe for every backend with an lmcp_url, via a single
+-- bash fan-out of curl calls. Total wall clock ≈ PROBE_BUDGET.
+local function probe_all_parallel(force)
+    local now = os.time()
+    local need = {}
+    for name, b in pairs(backends) do
+        if b.lmcp_url and (force or not cache_fresh(status[name], now)) then
+            need[#need+1] = b
         end
     end
-    if not new_s.up and b.ssh_host then
-        local _, _, rc = ssh_run_script(b.ssh_host, "exit 0\n")
-        if rc == 0 then
-            new_s.up = true
-            new_s.via = "ssh"
-            new_s.err = nil
-        else
-            new_s.err = (new_s.err or "") .. "; ssh: rc=" .. tostring(rc)
+    if #need == 0 then return end
+
+    local script_parts = {}
+    for _, b in ipairs(need) do
+        local auth = b.token and (" -H 'Authorization: Bearer " .. b.token .. "'") or ""
+        local url = b.lmcp_url:gsub("'", "'\\''")
+        script_parts[#script_parts+1] = string.format(
+            "(curl --max-time %d -s -o /dev/null -w '%s %%{http_code} %%{time_total}\\n' -X POST%s -H 'Content-Type: application/json' -d '{\"jsonrpc\":\"2.0\",\"id\":1,\"method\":\"tools/list\",\"params\":{}}' '%s' || echo '%s ERR 0') &",
+            PROBE_BUDGET, b.name, auth, url, b.name
+        )
+    end
+    script_parts[#script_parts+1] = "wait"
+
+    local t0 = monotonic()
+    local p = io.popen(table.concat(script_parts, "\n"))
+    local out = p and p:read("*a") or ""
+    if p then p:close() end
+    local dt = monotonic() - t0
+
+    local seen = {}
+    for line in out:gmatch("[^\n]+") do
+        local name, code, t = line:match("^(%S+)%s+(%S+)%s+([%d%.]+)")
+        if name then
+            seen[name] = true
+            local is_up = (code == "200")
+            if is_up then
+                apply_probe_result(name, true, nil, "lmcp", nil)
+            else
+                apply_probe_result(name, false, "lmcp code=" .. code, nil, nil)
+            end
         end
     end
-    status[name] = new_s
-    return new_s
+    -- Hosts with no output (fan-out error): mark DOWN
+    for _, b in ipairs(need) do
+        if not seen[b.name] then
+            apply_probe_result(b.name, false, "probe fan-out missing", nil, nil)
+        end
+    end
+    logreq("probe_all_parallel n=%d elapsed=%.2fs", #need, dt)
 end
 
 -- ---- Call-tool dispatcher ----------------------------------------------
@@ -221,16 +324,20 @@ end
 -- `allow_ssh`: whether the tool has an SSH fallback path
 -- `ssh_impl`: function(backend, args) -> output string, or nil + error
 local function call_remote(tool, args, allow_ssh, ssh_impl)
+    local t_start = monotonic()
     local host = args.host
     if type(host) ~= "string" or host == "" then
+        logreq("tool=%s ERR missing-host", tool)
         return "Error: missing `host` parameter"
     end
     local b = backends[host]
     if not b then
+        logreq("tool=%s host=%s ERR unknown-host", tool, host)
         return string.format("Error: unknown host %q (registry: %s)", host, CONF_PATH)
     end
 
     local errs = {}
+    local via = nil
 
     -- Try lmcp first
     if b.lmcp_url then
@@ -239,6 +346,8 @@ local function call_remote(tool, args, allow_ssh, ssh_impl)
         local result, err = jsonrpc_call(b.lmcp_url, b.token, "tools/call",
             { name = tool, arguments = args_pass })
         if result then
+            via = "lmcp"
+            logreq("tool=%s host=%s via=lmcp elapsed=%.2fs", tool, host, monotonic() - t_start)
             -- Propagate backend's content, extract first text block
             if type(result.content) == "table" and result.content[1] and result.content[1].text then
                 local prefix = result.isError and "[lmcp isError] " or ""
@@ -254,12 +363,16 @@ local function call_remote(tool, args, allow_ssh, ssh_impl)
     -- Fallback to ssh
     if allow_ssh and ssh_impl and b.ssh_host then
         local out, serr = ssh_impl(b, args)
-        if out ~= nil then return "[via ssh fallback]\n" .. out end
+        if out ~= nil then
+            logreq("tool=%s host=%s via=ssh elapsed=%.2fs", tool, host, monotonic() - t_start)
+            return "[via ssh fallback]\n" .. out
+        end
         errs[#errs+1] = "ssh: " .. tostring(serr)
     elseif allow_ssh and not b.ssh_host then
         errs[#errs+1] = "ssh: no host configured"
     end
 
+    logreq("tool=%s host=%s FAIL elapsed=%.2fs err=%s", tool, host, monotonic() - t_start, table.concat(errs, " | "):sub(1, 200))
     return "Error: " .. table.concat(errs, " | ")
 end
 
@@ -355,7 +468,7 @@ math.randomseed(os.time())
 
 local server = lmcp.new(os.getenv("LMCP_NAME") or "hub-tools", {
     port = tonumber(os.getenv("LMCP_PORT") or arg[1]) or 8090,
-    version = "0.5.0",
+    version = "0.5.4",
     conf = os.getenv("LMCP_HUB_CONF") or "/opt/herding/etc/lmcp-hub.conf",
 })
 
@@ -371,13 +484,14 @@ server:tool("remote_list_hosts",
     } },
     function(a)
         local force = a and a.force or false
+        probe_all_parallel(force)
         local lines = {}
         local names = {}
         for n in pairs(backends) do names[#names+1] = n end
         table.sort(names)
         for _, n in ipairs(names) do
             local b = backends[n]
-            local s = probe(n, force)
+            local s = status[n] or { up = false, err = "no probe result" }
             local paths = {}
             if b.lmcp_url then paths[#paths+1] = "lmcp" end
             if b.ssh_host then paths[#paths+1] = "ssh" end
@@ -391,7 +505,14 @@ server:tool("remote_list_hosts",
             )
         end
         return table.concat(lines, "\n")
-    end
+    end,
+    { annotations = {
+        title           = "List fleet hosts",
+        readOnlyHint    = true,
+        destructiveHint = false,
+        idempotentHint  = true,
+        openWorldHint   = true,
+    } }
 )
 
 server:tool("remote_shell", "Run a shell command on a fleet host. lmcp-primary with ssh fallback.",
@@ -401,7 +522,14 @@ server:tool("remote_shell", "Run a shell command on a fleet host. lmcp-primary w
         cwd = { type = "string", description = "Working directory" },
         timeout = { type = "integer", description = "Timeout (seconds)", default = 120 },
     }, required = { "host", "command" } },
-    function(a) return call_remote("shell", a, true, ssh_shell) end
+    function(a) return call_remote("shell", a, true, ssh_shell) end,
+    { annotations = {
+        title           = "Remote shell",
+        readOnlyHint    = false,
+        destructiveHint = true,
+        idempotentHint  = false,
+        openWorldHint   = true,
+    } }
 )
 
 server:tool("remote_read_file", "Read a file from a fleet host.",
@@ -409,7 +537,14 @@ server:tool("remote_read_file", "Read a file from a fleet host.",
         host = HOST_ARG,
         path = { type = "string", description = "File path" },
     }, required = { "host", "path" } },
-    function(a) return call_remote("read_file", a, true, ssh_read_file) end
+    function(a) return call_remote("read_file", a, true, ssh_read_file) end,
+    { annotations = {
+        title           = "Remote read file",
+        readOnlyHint    = true,
+        destructiveHint = false,
+        idempotentHint  = true,
+        openWorldHint   = true,
+    } }
 )
 
 server:tool("remote_write_file", "Write content to a file on a fleet host.",
@@ -418,7 +553,14 @@ server:tool("remote_write_file", "Write content to a file on a fleet host.",
         path = { type = "string" },
         content = { type = "string" },
     }, required = { "host", "path", "content" } },
-    function(a) return call_remote("write_file", a, true, ssh_write_file) end
+    function(a) return call_remote("write_file", a, true, ssh_write_file) end,
+    { annotations = {
+        title           = "Remote write file",
+        readOnlyHint    = false,
+        destructiveHint = true,
+        idempotentHint  = true,
+        openWorldHint   = true,
+    } }
 )
 
 server:tool("remote_edit_file",
@@ -430,7 +572,32 @@ server:tool("remote_edit_file",
         new_string = { type = "string" },
         replace_all = { type = "boolean", default = false },
     }, required = { "host", "path", "old_string", "new_string" } },
-    function(a) return call_remote("edit_file", a, false, nil) end
+    function(a) return call_remote("edit_file", a, false, nil) end,
+    { annotations = {
+        title           = "Remote edit file",
+        readOnlyHint    = false,
+        destructiveHint = true,
+        idempotentHint  = false,
+        openWorldHint   = true,
+    } }
+)
+
+server:tool("remote_shell_bg",
+    "Launch a detached background command on a fleet host (Linux). Returns PID + log path immediately. Requires backend lmcp v0.5.2+.",
+    { type = "object", properties = {
+        host = HOST_ARG,
+        command = { type = "string", description = "Shell command" },
+        cwd = { type = "string" },
+        log = { type = "string", description = "Log file path" },
+    }, required = { "host", "command" } },
+    function(a) return call_remote("shell_bg", a, false, nil) end,
+    { annotations = {
+        title           = "Remote shell (background)",
+        readOnlyHint    = false,
+        destructiveHint = true,
+        idempotentHint  = false,
+        openWorldHint   = true,
+    } }
 )
 
 server:tool("remote_list_dir", "List directory entries on a fleet host.",
@@ -438,7 +605,14 @@ server:tool("remote_list_dir", "List directory entries on a fleet host.",
         host = HOST_ARG,
         path = { type = "string", default = "." },
     }, required = { "host" } },
-    function(a) return call_remote("list_dir", a, true, ssh_list_dir) end
+    function(a) return call_remote("list_dir", a, true, ssh_list_dir) end,
+    { annotations = {
+        title           = "Remote list directory",
+        readOnlyHint    = true,
+        destructiveHint = false,
+        idempotentHint  = true,
+        openWorldHint   = true,
+    } }
 )
 
 server:tool("remote_search_files", "find-by-pattern on a fleet host.",
@@ -447,7 +621,14 @@ server:tool("remote_search_files", "find-by-pattern on a fleet host.",
         pattern = { type = "string" },
         path = { type = "string", default = "/" },
     }, required = { "host", "pattern" } },
-    function(a) return call_remote("search_files", a, true, ssh_search_files) end
+    function(a) return call_remote("search_files", a, true, ssh_search_files) end,
+    { annotations = {
+        title           = "Remote find files",
+        readOnlyHint    = true,
+        destructiveHint = false,
+        idempotentHint  = true,
+        openWorldHint   = true,
+    } }
 )
 
 io.stderr:write(string.format("lmcp-hub starting on port %d with %d backends from %s\n",

json.lua

@@ -60,6 +60,13 @@ encode_value = function(v)
     local t = type(v)
     if v == nil or v == json.null then
         return 'null'
+    elseif v == json.empty_object then
+        -- Sentinel for forcing {} (object) instead of [] (array) when
+        -- the field semantically requires an object but is empty.
+        -- Without this, every empty Lua table goes through is_array()
+        -- and emits as [], breaking spec-strict JSON-RPC consumers
+        -- (e.g. ping result, MUST be {}).
+        return '{}'
     elseif t == 'boolean' then
         return v and 'true' or 'false'
     elseif t == 'number' then
@@ -110,9 +117,20 @@ local function decode_string(s, pos)
             pos = pos + 1
             c = s:sub(pos, pos)
             if c == 'u' then
-                local hex = s:sub(pos + 1, pos + 4)
-                parts[#parts + 1] = utf8.char(tonumber(hex, 16))
+                local cp = tonumber(s:sub(pos + 1, pos + 4), 16)
                 pos = pos + 5
+                -- Combine UTF-16 surrogate pair so non-BMP chars (emoji,
+                -- supplementary CJK) decode correctly instead of as two
+                -- lone surrogates → invalid UTF-8.
+                if cp and cp >= 0xD800 and cp <= 0xDBFF
+                   and s:sub(pos, pos + 1) == "\\u" then
+                    local lo = tonumber(s:sub(pos + 2, pos + 5), 16)
+                    if lo and lo >= 0xDC00 and lo <= 0xDFFF then
+                        cp = (cp - 0xD800) * 0x400 + (lo - 0xDC00) + 0x10000
+                        pos = pos + 6
+                    end
+                end
+                parts[#parts + 1] = utf8.char(cp)
             else
                 local esc = { n = '\n', r = '\r', t = '\t', b = '\b', f = '\f' }
                 parts[#parts + 1] = esc[c] or c
@@ -210,6 +228,12 @@ end
 -- Sentinel for JSON null
 json.null = setmetatable({}, { __tostring = function() return 'null' end })
 
+-- Sentinel for an empty JSON object ({}). Use when a field semantically
+-- requires an object but is empty — e.g. `ping` result, MCP _meta = {}.
+-- Without this, an empty Lua table goes through is_array() → '[]'.
+-- See memory project_json_empty_table_gotcha.md.
+json.empty_object = setmetatable({}, { __tostring = function() return '{}' end })
+
 -- Helper: encode a table as a JSON array even if empty
 function json.array(t)
     return setmetatable(t or {}, { __is_array = true })

tools.d/hertz.lua

@@ -0,0 +1,292 @@
+-- /opt/lmcp/tools.d/hertz.lua — hertz-specific tool registrations.
+--
+-- Invoked by the packaged server.lua's tools.d scan (lmcp ≥ v1.2.0, issue #22).
+-- Receives (server, run) — the configured lmcp instance and the
+-- coroutine-aware run() helper. We add hertz-only tools on top of the
+-- packaged generics (shell, read_file, write_file, edit_file, list_dir,
+-- search_files, fetch, web_search, shell_bg).
+--
+-- Previously these all lived in /opt/lmcp/server.lua (a copy-paste fork of
+-- the packaged server.lua). That pattern drifted on every release; now the
+-- packaged file stays canonical and only the genuine hertz-specifics live
+-- here.
+
+local server, run = ...
+
+-- Local helpers — single-host scoped, not worth lifting into the packaged lib.
+local function read_file_raw(path)
+    local f = io.open(path, 'r')
+    if not f then return nil end
+    local c = f:read('*a'); f:close(); return c
+end
+
+local function farad(cmd, timeout)
+    return run("incus exec farad -- sh -c " .. string.format("%q", cmd),
+        timeout or 15)
+end
+
+-- ---- LXD tools ----
+
+server:tool("incus_exec", "Execute a command inside an Incus container on hertz.", {
+    type = "object",
+    properties = {
+        container = { type = "string", description = "Container name" },
+        command   = { type = "string", description = "Command to execute" },
+        timeout   = { type = "integer", default = 30 },
+    },
+    required = { "container", "command" },
+}, function(a)
+    return run(string.format("incus exec %s -- sh -c %q",
+        a.container:gsub("[^%w%-]", ""), a.command), a.timeout or 30)
+end)
+
+server:tool("incus_list", "List all Incus containers on hertz.",
+    { type = "object" },
+    function() return run("incus list -c ns4 -f csv", 10) end)
+
+-- ---- Fritz!Box tools ----
+
+server:tool("fritz", "Execute Fritz!Box TR-064 command (info, hosts, wan, "
+    .. "wol <mac>, reconnect, reboot, reboot-repeater <ip>, routes, route-add, "
+    .. "route-del, services, actions, call).",
+    {
+        type = "object",
+        properties = {
+            command = { type = "string", description = "fritz subcommand and args" },
+        },
+        required = { "command" },
+    },
+    function(a) return run("sudo /root/.local/bin/fritz " .. a.command, 15) end)
+
+-- ---- Network tools ----
+
+server:tool("ping_host", "Check if a host is reachable (1 ICMP ping, 2s timeout).", {
+    type = "object",
+    properties = { host = { type = "string" } },
+    required = { "host" },
+}, function(a)
+    local host = a.host:gsub("[^%w%.%-:]", "")
+    return run("ping -c1 -W2 " .. host, 5)
+end)
+
+server:tool("network_status",
+    "Check reachability of all infrastructure hosts and MCP endpoints.",
+    { type = "object" },
+    function()
+        local script = [[
+hosts="hertz:localhost boltzmann:boltzmann tesla:tesla data:192.168.88.30 broglie:192.168.88.160 higgs:10.170.16.10"
+for entry in $hosts; do
+    name="${entry%%:*}"
+    ip="${entry#*:}"
+    if ping -c1 -W2 "$ip" >/dev/null 2>&1; then
+        status="UP"
+        if [ "$name" != "data" ]; then
+            if timeout 3 bash -c "echo >/dev/tcp/${ip}/8080" 2>/dev/null; then
+                status="UP (MCP ok)"
+            else
+                status="UP (no MCP)"
+            fi
+        fi
+    else
+        status="DOWN"
+    fi
+    printf "%-12s %-20s %s\n" "$name" "$ip" "$status"
+done
+]]
+        return run(script, 30)
+    end)
+
+server:tool("wol_and_wait",
+    "Wake the data server via Fritz!Box WoL and wait until it (or broglie MCP) is reachable.",
+    {
+        type = "object",
+        properties = {
+            mac          = { type = "string",  default = "", description = "MAC address (auto-detected if empty)" },
+            wait_for_mcp = { type = "boolean", default = true, description = "Wait for broglie MCP instead of just ping" },
+            timeout      = { type = "integer", default = 120 },
+        },
+    },
+    function(a)
+        local mac = a.mac
+        if not mac or mac == "" then
+            local hosts_out = run("sudo /root/.local/bin/fritz hosts 2>/dev/null | grep -i 'data\\|192.168.88.30'", 10)
+            mac = hosts_out and hosts_out:match("(%x%x:%x%x:%x%x:%x%x:%x%x:%x%x)") or ""
+            if mac == "" then return "Error: could not detect MAC for data. Provide it manually." end
+        end
+        run("sudo /root/.local/bin/fritz wol " .. mac:gsub("[^%x:]", ""), 10)
+
+        local timeout = a.timeout or 120
+        local check
+        if a.wait_for_mcp == false then
+            check = "ping -c1 -W2 192.168.88.30"
+        else
+            check = "timeout 3 bash -c 'echo >/dev/tcp/192.168.88.160/8080'"
+        end
+        local target = (a.wait_for_mcp == false) and "data" or "broglie:8080"
+
+        local elapsed = 0
+        while elapsed < timeout do
+            os.execute("sleep 5")
+            elapsed = elapsed + 5
+            local rc = os.execute(check)
+            if rc == true or rc == 0 then
+                return string.format("WoL sent to %s. %s reachable after %ds.",
+                    mac, target, elapsed)
+            end
+        end
+        return string.format("WoL sent to %s but %s not reachable after %ds.",
+            mac, target, timeout)
+    end)
+
+-- ---- Proxmox fallback ----
+
+server:tool("pct_exec",
+    "Execute a command in a Proxmox CT on data (SSH fallback when broglie is down).",
+    {
+        type = "object",
+        properties = {
+            ctid    = { type = "string", description = "Container ID (e.g. 108, 165)" },
+            command = { type = "string" },
+            timeout = { type = "integer", default = 30 },
+        },
+        required = { "ctid", "command" },
+    },
+    function(a)
+        local ctid = a.ctid:gsub("[^%d]", "")
+        return run(string.format("ssh -o ConnectTimeout=5 root@data 'pct exec %s -- sh -c %q'",
+            ctid, a.command), a.timeout or 30)
+    end)
+
+-- ---- Home Assistant ----
+
+server:tool("ha_cli",
+    "Run Home Assistant CLI command inside farad (e.g. 'core restart', 'backups list', 'info').",
+    {
+        type = "object",
+        properties = {
+            command  = { type = "string", description = "ha subcommand, e.g. 'core restart', 'supervisor info', 'backups list'" },
+            raw_json = { type = "boolean", default = false, description = "Return raw JSON output" },
+        },
+        required = { "command" },
+    },
+    function(a)
+        local flags = a.raw_json and " --raw-json" or ""
+        return farad("ha " .. a.command .. flags)
+    end)
+
+server:tool("ha_api",
+    "Call Home Assistant Core REST API. Requires token in /opt/lmcp/ha_token on hertz.",
+    {
+        type = "object",
+        properties = {
+            method   = { type = "string", default = "GET" },
+            endpoint = { type = "string", description = "/api/states, /api/services/climate/set_temperature, etc." },
+            body     = { type = "string", description = "JSON body for POST" },
+        },
+        required = { "endpoint" },
+    },
+    function(a)
+        local token = read_file_raw("/opt/lmcp/ha_token")
+        if not token or token == "" then
+            return "Error: no HA token. Create a long-lived token in HA UI → Profile → "
+                .. "Long-Lived Access Tokens, then: echo '<token>' | sudo tee /opt/lmcp/ha_token"
+        end
+        token = token:gsub("%s+", "")
+        local method = (a.method or "GET"):upper()
+        local cmd = string.format(
+            "curl -sf -X %s -H 'Authorization: Bearer %s' -H 'Content-Type: application/json'",
+            method, token)
+        if a.body and a.body ~= "" then
+            cmd = cmd .. " -d " .. string.format("'%s'", a.body:gsub("'", "'\\''"))
+        end
+        cmd = cmd .. " http://localhost:8123" .. a.endpoint
+        return farad(cmd, 15)
+    end)
+
+-- ---- MQTT (Eurotronic thermostats + general) ----
+
+server:tool("mqtt_pub", "Publish an MQTT message (via Mosquitto in farad).", {
+    type = "object",
+    properties = {
+        topic   = { type = "string" },
+        message = { type = "string" },
+        retain  = { type = "boolean", default = false },
+    },
+    required = { "topic", "message" },
+}, function(a)
+    local retain = a.retain and "-r " or ""
+    return farad(string.format("mosquitto_pub %s-t '%s' -m '%s'",
+        retain, a.topic:gsub("'", "'\\''"), a.message:gsub("'", "'\\''")))
+end)
+
+server:tool("mqtt_sub",
+    "Subscribe to MQTT topics and collect messages (via Mosquitto in farad).",
+    {
+        type = "object",
+        properties = {
+            topic   = { type = "string",  description = "Topic pattern, e.g. 'eurotronic/#' or '#'" },
+            count   = { type = "integer", default = 10, description = "Number of messages to collect" },
+            timeout = { type = "integer", default = 5,  description = "Seconds to wait" },
+            verbose = { type = "boolean", default = true, description = "Show topics with messages" },
+        },
+        required = { "topic" },
+    },
+    function(a)
+        local v = a.verbose ~= false and "-v " or ""
+        return farad(string.format("mosquitto_sub %s-t '%s' -C %d -W %d",
+            v, a.topic:gsub("'", "'\\''"), a.count or 10, a.timeout or 5),
+            (a.timeout or 5) + 5)
+    end)
+
+-- ---- Mediagrab (kids show downloader) ----
+
+server:tool("mediagrab",
+    "Manage kids show downloads in doppler container. Commands: list, weekly, "
+    .. "archive, test <url>, add '<json>'",
+    {
+        type = "object",
+        properties = {
+            command = { type = "string", description = "Command: list, weekly, archive, 'test <url>', 'add <json>'" },
+        },
+        required = { "command" },
+    },
+    function(a)
+        return run("incus exec doppler -- python3 /opt/mediagrab/mediagrab.py "
+            .. a.command, 120)
+    end)
+
+-- ---- Fleet wake (pipi: wake-only, no power-off) ----
+server:tool("wake_fleet",
+    "Wake a fleet NUC (pve1..pve4) via Fritz!Box Wake-on-LAN. Powers a node ON only; it cannot power anything off. Node boots in ~30-60s.",
+    {
+        type = "object",
+        properties = {
+            node = { type = "string", description = "Node to wake: '1'..'4' or 'pve1'..'pve4'" },
+        },
+        required = { "node" },
+    },
+    function(a)
+        local node = tostring(a.node or ""):gsub("[^%w]", "")
+        if not node:match("^p?v?e?[1-4]$") then
+            return "Error: node must be 1-4 or pve1-pve4 (got: " .. tostring(a.node) .. ")"
+        end
+        return run("sudo /root/.local/bin/wake-pve " .. node .. " 2>&1", 15)
+    end)
+
+-- ---- apropos: lean facade for stash recall (read-only memory) ----
+server:tool("apropos",
+    "Search shared fleet memory (stash) for facts about the fleet, projects, decisions, and preferences. query = 2-6 words on the topic; limit = max results (default 3). Read-only.",
+    {
+        type = "object",
+        properties = {
+            query = { type = "string",  description = "2-6 words describing what to recall" },
+            limit = { type = "integer", description = "max results, default 3" },
+        },
+        required = { "query" },
+    },
+    function(a)
+        local q = tostring(a.query or ""):gsub("[^%w%s%-%.]", " "):gsub("%s+", " ")
+        if q:gsub("%s","") == "" then return "Error: query required" end
+        local lim = tonumber(a.limit) or 3
+        return run("python3 /opt/lmcp/helpers/stash_recall.py '" .. q .. "' " .. lim, 30)
+    end)

tools.d/nash.lua

@@ -0,0 +1,103 @@
+-- /opt/lmcp/tools.d/nash.lua -- nash shared memory tools
+--
+-- Provides add/search/list/delete tools for the nash memory service.
+-- Requires NASH_URL env var (e.g. http://192.168.88.143:8000).
+-- When unset, tools return an error guiding the user to set it.
+--
+-- SPDX-License-Identifier: MIT
+
+local server, run = ...
+
+local function nash_url()
+    local url = os.getenv("NASH_URL")
+    if not url or url == "" then return nil end
+    return url:gsub("/+$", "")
+end
+
+local function curl_json(url, method, body, timeout)
+    timeout = timeout or 10
+    local out = "/tmp/lmcp-nash-" .. os.time() .. "-" .. math.random(10000, 99999) .. ".json"
+    local fmt = "http_code=%{http_code}"
+    local cmd
+    if body then
+        local tmp = out .. ".req"
+        local f = io.open(tmp, "w")
+        if f then f:write(body); f:close() end
+        cmd = string.format(
+            "curl -sS -X %s --max-time %d -H 'Content-Type: application/json' --data-binary '@%s' -o '%s' -w '%s' '%s'",
+            method, timeout, tmp, out, fmt, url
+        )
+        os.execute("rm -f '" .. tmp .. "' 2>/dev/null")
+    else
+        cmd = string.format(
+            "curl -sS -X %s --max-time %d -o '%s' -w '%s' '%s'",
+            method, timeout, out, fmt, url
+        )
+    end
+    local raw = run(cmd, timeout + 5) or ""
+    local http_code = tonumber(raw:match("(%d+)$")) or 0
+    local body_out = ""
+    local f = io.open(out, "r")
+    if f then body_out = f:read("*a") or ""; f:close() end
+    os.execute("rm -f '" .. out .. "' 2>/dev/null")
+    return body_out, http_code
+end
+
+server:tool("nash_add", "Store a text entry in shared nash memory.", {
+    type = "object",
+    properties = {
+        text = { type = "string", description = "Text content to remember" },
+    },
+    required = { "text" },
+}, function(a)
+    local base = nash_url()
+    if not base then return "Error: set NASH_URL env var (e.g. http://192.168.88.143:8000)" end
+    local json, code = curl_json(base .. "/add", "POST", '{"text":' .. require("json").encode(a.text) .. '}')
+    if code ~= 200 then return "Error: nash API HTTP " .. code .. ": " .. json end
+    return json
+end)
+
+server:tool("nash_search", "Semantic search across shared nash memory.", {
+    type = "object",
+    properties = {
+        query = { type = "string", description = "Search query" },
+        limit = { type = "integer", description = "Max results (default 5)", default = 5 },
+    },
+    required = { "query" },
+}, function(a)
+    local base = nash_url()
+    if not base then return "Error: set NASH_URL env var (e.g. http://192.168.88.143:8000)" end
+    local body = '{"query":' .. require("json").encode(a.query) .. ',"limit":' .. (a.limit or 5) .. '}'
+    local json, code = curl_json(base .. "/search", "POST", body)
+    if code ~= 200 then return "Error: nash API HTTP " .. code .. ": " .. json end
+    return json
+end)
+
+server:tool("nash_list", "List all entries in shared nash memory.", {
+    type = "object",
+    properties = {
+        limit = { type = "integer", description = "Max results (default 100)", default = 100 },
+    },
+}, function(a)
+    local base = nash_url()
+    if not base then return "Error: set NASH_URL env var (e.g. http://192.168.88.143:8000)" end
+    local body = '{"limit":' .. (a.limit or 100) .. '}'
+    local json, code = curl_json(base .. "/scroll", "POST", body)
+    if code ~= 200 then return "Error: nash API HTTP " .. code .. ": " .. json end
+    return json
+end)
+
+server:tool("nash_delete", "Delete an entry from shared nash memory by ID.", {
+    type = "object",
+    properties = {
+        id = { type = "string", description = "Entry ID to delete" },
+    },
+    required = { "id" },
+}, function(a)
+    local base = nash_url()
+    if not base then return "Error: set NASH_URL env var (e.g. http://192.168.88.143:8000)" end
+    local body = '{"id":' .. require("json").encode(a.id) .. '}'
+    local json, code = curl_json(base .. "/delete", "POST", body)
+    if code ~= 200 then return "Error: nash API HTTP " .. code .. ": " .. json end
+    return json
+end)

windows/README.md

@@ -0,0 +1,57 @@
+# lmcp Windows MSI build
+
+This directory contains the WiX manifest and packaging files for the
+Windows MSI build of lmcp.
+
+## Recommended: cross-build on Linux (one command)
+
+```sh
+./build-msi.sh /path/to/output/dir
+```
+
+Downloads Lua 5.4 Win64 binaries from LuaBinaries, cross-compiles
+LuaSocket via `mingw-w64`, stages `pkg/lua/`, and runs `wixl` to
+produce `lmcp-<version>.msi`. No Windows VM required.
+
+Prereqs on a Debian/Ubuntu builder:
+```sh
+sudo apt install wixl unzip gcc-mingw-w64-x86-64 \
+                 binutils-mingw-w64-x86-64 mingw-w64-x86-64-dev curl
+```
+
+Version comes from `lmcp.wxs` `Version="…"`. Bump that before
+building a release.
+
+## Alternative: build on Windows via WiX toolset
+
+```cmd
+sync.sh                                  REM see "tracked vs. generated"
+REM ensure pkg/lua/ has the runtime — see below
+candle.exe lmcp.wxs
+light.exe lmcp.wixobj -o lmcp-1.x.y.msi
+```
+
+## What's tracked vs. generated
+
+- **Tracked** (edit in git):
+  - `lmcp.wxs` — WiX MSI manifest
+  - `sync.sh` — copies root .lua sources → `pkg/`
+  - `README.md` — this file
+  - `pkg/install_service.bat` — Windows service installer
+  - `pkg/start.bat` — manual launcher
+
+- **Generated / external** (gitignored):
+  - `pkg/lmcp.lua`, `pkg/server.lua`, `pkg/json.lua` — produced by
+    `sync.sh`. Never edit directly; edit the root files and re-sync.
+  - `pkg/lua/` — the Lua + LuaSocket runtime drop-in. Download
+    separately and place here. Suggested source: the lua-binaries
+    project (https://github.com/rjpcomputing/luaforwindows) or a
+    similar pre-built bundle. The MSI expects `pkg/lua/lua.exe`,
+    `pkg/lua/lua54.dll`, and the `pkg/lua/socket/` + `pkg/lua/mime/`
+    subdirectories per the manifest.
+
+## Issue history
+
+Issue #18 (closed in v1.1.0) introduced this workflow after the
+`pkg/` lua sources had silently drifted ~6 months out of date,
+missing every feature added since April 2026.

windows/build-msi.sh

@@ -0,0 +1,100 @@
+#!/bin/sh
+# windows/build-msi.sh — produce lmcp-<ver>.msi on Linux via wixl.
+#
+# This is the first-time-discovered cross-build path: download Lua 5.4
+# Win64 binaries from LuaBinaries, cross-compile LuaSocket with mingw-w64,
+# stage windows/pkg/lua/, then invoke wixl on the WiX manifest.
+#
+# Avoids the VM106-clone + WiX-on-Windows path entirely. ~1 minute on a
+# warm cache; ~3-5 minutes cold (downloads ~700 KB + cross-compiles).
+#
+# Prereqs (apt install on Debian aarch64):
+#   apt-get install -y wixl unzip gcc-mingw-w64-x86-64 binutils-mingw-w64-x86-64 \
+#                      mingw-w64-x86-64-dev
+#
+# Usage: ./build-msi.sh [output_dir]
+#   Output: $output_dir/lmcp-<ver>.msi (default: $PWD)
+#
+# Version comes from windows/lmcp.wxs Version="…" attribute.
+set -eu
+
+here=$(dirname "$(readlink -f "$0")")
+root=$(cd "$here/.." && pwd)
+out_dir=${1:-$PWD}
+work=$(mktemp -d /tmp/lmcp-msi-XXXXXX)
+trap "rm -rf $work" EXIT
+
+# Versions — bump as upstream releases.
+LUA_VER=5.4.2
+LUASOCKET_VER=3.1.0
+
+# Pull current lmcp version from the WiX manifest.
+lmcp_ver=$(sed -n 's/.*Version="\([^"]*\)".*/\1/p' "$here/lmcp.wxs" | head -1)
+[ -n "$lmcp_ver" ] || { echo "build-msi.sh: cannot parse Version from lmcp.wxs" >&2; exit 1; }
+
+echo "build-msi.sh: lmcp $lmcp_ver, lua $LUA_VER, luasocket $LUASOCKET_VER"
+
+echo "==> 1/5 sync lmcp .lua sources into pkg/"
+"$here/sync.sh"
+
+echo "==> 2/5 fetch lua $LUA_VER win64 binaries + dev library"
+cd "$work"
+curl -sSLf -o lua-bin.zip \
+    "https://downloads.sourceforge.net/project/luabinaries/${LUA_VER}/Tools%20Executables/lua-${LUA_VER}_Win64_bin.zip"
+curl -sSLf -o lua-lib.zip \
+    "https://downloads.sourceforge.net/project/luabinaries/${LUA_VER}/Windows%20Libraries/Dynamic/lua-${LUA_VER}_Win64_dllw6_lib.zip"
+mkdir -p luabin lualib include/lua/54 include/lua54 bin/lua/54 bin/lua54 lib/lua/54 lib/lua54
+unzip -q -o lua-bin.zip -d luabin
+unzip -q -o lua-lib.zip -d lualib
+cp lualib/include/*.h include/lua/54/
+cp lualib/include/*.h include/lua54/
+cp lualib/liblua54.a lib/lua/54/
+cp lualib/liblua54.a lib/lua54/
+cp lualib/lua54.dll bin/lua/54/
+cp lualib/lua54.dll bin/lua54/
+
+echo "==> 3/5 cross-compile LuaSocket $LUASOCKET_VER for win64"
+curl -sSLf -o luasocket.tar.gz \
+    "https://github.com/lunarmodules/luasocket/archive/refs/tags/v${LUASOCKET_VER}.tar.gz"
+tar xzf luasocket.tar.gz
+cd "luasocket-${LUASOCKET_VER}"
+make -s PLAT=mingw \
+    CC=x86_64-w64-mingw32-gcc \
+    LD=x86_64-w64-mingw32-gcc \
+    LUAV=54 \
+    LUAINC_mingw_base="$work/include" \
+    LUALIB_mingw_base="$work/bin" \
+    > /dev/null
+
+echo "==> 4/5 stage pkg/lua/"
+pkg_lua="$here/pkg/lua"
+rm -rf "$pkg_lua"
+mkdir -p "$pkg_lua/socket" "$pkg_lua/mime"
+
+# WiX manifest expects "lua.exe" (not "lua54.exe").
+cp "$work/luabin/lua54.exe"        "$pkg_lua/lua.exe"
+cp "$work/luabin/lua54.dll"        "$pkg_lua/lua54.dll"
+cp src/socket.lua                  "$pkg_lua/"
+cp src/mime.lua                    "$pkg_lua/"
+cp src/ltn12.lua                   "$pkg_lua/"
+cp src/socket-3.0.0.dll            "$pkg_lua/socket/core.dll"
+cp src/ftp.lua                     "$pkg_lua/socket/"
+cp src/headers.lua                 "$pkg_lua/socket/"
+cp src/http.lua                    "$pkg_lua/socket/"
+cp src/smtp.lua                    "$pkg_lua/socket/"
+cp src/tp.lua                      "$pkg_lua/socket/"
+cp src/url.lua                     "$pkg_lua/socket/"
+cp src/mime-1.0.3.dll              "$pkg_lua/mime/core.dll"
+
+echo "==> 5/5 wixl: produce MSI"
+# wixl wants forward slashes; rewrite Windows-style backslashes in Source=.
+wxs_tmp="$work/lmcp.wxs"
+sed 's|Source="pkg\\|Source="pkg/|g; s|\\\([a-zA-Z]\)|/\1|g' "$here/lmcp.wxs" > "$wxs_tmp"
+mkdir -p "$out_dir"
+out_msi="$out_dir/lmcp-${lmcp_ver}.msi"
+(cd "$here" && wixl -v "$wxs_tmp" -o "$out_msi")
+
+echo ""
+echo "==> done: $out_msi"
+ls -la "$out_msi"
+sha256sum "$out_msi"

windows/lmcp.wxs

@@ -0,0 +1,119 @@
+<?xml version="1.0" encoding="utf-8"?>
+<Wix xmlns="http://schemas.microsoft.com/wix/2006/wi">
+  <!-- Bump Version on every release. See windows/README.md. -->
+  <Product Id="*"
+           Name="lmcp — Lua MCP Server"
+           Language="1033"
+           Version="1.1.0"
+           Manufacturer="QAP'LA Project"
+           UpgradeCode="A7F3E2D1-4B5C-6D7E-8F9A-0B1C2D3E4F5A">
+
+    <Package InstallerVersion="200"
+             Compressed="yes"
+             InstallScope="perMachine"
+             Description="Lightweight MCP server in Lua. 2MB RSS."
+             Comments="Zero-dependency MCP server." />
+
+    <MediaTemplate EmbedCab="yes" />
+
+    <MajorUpgrade DowngradeErrorMessage="A newer version is already installed." />
+
+    <Directory Id="TARGETDIR" Name="SourceDir">
+      <Directory Id="ProgramFiles64Folder">
+        <Directory Id="INSTALLFOLDER" Name="lmcp">
+          <Directory Id="LUA_DIR" Name="lua">
+            <Directory Id="SOCKET_DIR" Name="socket" />
+            <Directory Id="MIME_DIR" Name="mime" />
+          </Directory>
+        </Directory>
+      </Directory>
+    </Directory>
+
+    <!-- lmcp application files -->
+    <DirectoryRef Id="INSTALLFOLDER">
+      <Component Id="JsonLua" Guid="B1A2C3D4-E5F6-7890-ABCD-EF1234567890">
+        <File Id="json.lua" Source="pkg\json.lua" KeyPath="yes" />
+      </Component>
+      <Component Id="LmcpLua" Guid="B1A2C3D4-E5F6-7890-ABCD-EF1234567891">
+        <File Id="lmcp.lua" Source="pkg\lmcp.lua" KeyPath="yes" />
+      </Component>
+      <Component Id="ServerLua" Guid="B1A2C3D4-E5F6-7890-ABCD-EF1234567892">
+        <File Id="server.lua" Source="pkg\server.lua" KeyPath="yes" />
+      </Component>
+      <Component Id="StartBat" Guid="B1A2C3D4-E5F6-7890-ABCD-EF1234567893">
+        <File Id="start.bat" Source="pkg\start.bat" KeyPath="yes" />
+      </Component>
+      <Component Id="InstallService" Guid="B1A2C3D4-E5F6-7890-ABCD-EF1234567894">
+        <File Id="install_service.bat" Source="pkg\install_service.bat" KeyPath="yes" />
+      </Component>
+    </DirectoryRef>
+
+    <!-- Lua runtime -->
+    <DirectoryRef Id="LUA_DIR">
+      <Component Id="LuaExe" Guid="C2B3D4E5-F6A7-8901-BCDE-F12345678900">
+        <File Id="lua.exe" Source="pkg\lua\lua.exe" KeyPath="yes" />
+      </Component>
+      <Component Id="LuaDll" Guid="C2B3D4E5-F6A7-8901-BCDE-F12345678901">
+        <File Id="lua54.dll" Source="pkg\lua\lua54.dll" KeyPath="yes" />
+      </Component>
+      <Component Id="SocketLua" Guid="C2B3D4E5-F6A7-8901-BCDE-F12345678902">
+        <File Id="socket.lua" Source="pkg\lua\socket.lua" KeyPath="yes" />
+      </Component>
+      <Component Id="MimeLua" Guid="C2B3D4E5-F6A7-8901-BCDE-F12345678903">
+        <File Id="mime.lua" Source="pkg\lua\mime.lua" KeyPath="yes" />
+      </Component>
+      <Component Id="Ltn12Lua" Guid="C2B3D4E5-F6A7-8901-BCDE-F12345678904">
+        <File Id="ltn12.lua" Source="pkg\lua\ltn12.lua" KeyPath="yes" />
+      </Component>
+    </DirectoryRef>
+
+    <!-- LuaSocket native DLLs -->
+    <DirectoryRef Id="SOCKET_DIR">
+      <Component Id="SocketCoreDll" Guid="D3C4E5F6-A7B8-9012-CDEF-123456789010">
+        <File Id="socket_core.dll" Name="core.dll" Source="pkg\lua\socket\core.dll" KeyPath="yes" />
+      </Component>
+      <Component Id="SocketFtp" Guid="D3C4E5F6-A7B8-9012-CDEF-123456789011">
+        <File Id="ftp.lua" Source="pkg\lua\socket\ftp.lua" KeyPath="yes" />
+      </Component>
+      <Component Id="SocketHeaders" Guid="D3C4E5F6-A7B8-9012-CDEF-123456789012">
+        <File Id="headers.lua" Source="pkg\lua\socket\headers.lua" KeyPath="yes" />
+      </Component>
+      <Component Id="SocketHttp" Guid="D3C4E5F6-A7B8-9012-CDEF-123456789013">
+        <File Id="http.lua" Source="pkg\lua\socket\http.lua" KeyPath="yes" />
+      </Component>
+      <Component Id="SocketTp" Guid="D3C4E5F6-A7B8-9012-CDEF-123456789014">
+        <File Id="tp.lua" Source="pkg\lua\socket\tp.lua" KeyPath="yes" />
+      </Component>
+      <Component Id="SocketUrl" Guid="D3C4E5F6-A7B8-9012-CDEF-123456789015">
+        <File Id="url.lua" Source="pkg\lua\socket\url.lua" KeyPath="yes" />
+      </Component>
+    </DirectoryRef>
+
+    <DirectoryRef Id="MIME_DIR">
+      <Component Id="MimeCoreDll" Guid="E4D5F6A7-B8C9-0123-DEFA-234567890120">
+        <File Id="mime_core.dll" Name="core.dll" Source="pkg\lua\mime\core.dll" KeyPath="yes" />
+      </Component>
+    </DirectoryRef>
+
+    <Feature Id="MainFeature" Title="lmcp Server" Level="1">
+      <ComponentRef Id="JsonLua" />
+      <ComponentRef Id="LmcpLua" />
+      <ComponentRef Id="ServerLua" />
+      <ComponentRef Id="StartBat" />
+      <ComponentRef Id="InstallService" />
+      <ComponentRef Id="LuaExe" />
+      <ComponentRef Id="LuaDll" />
+      <ComponentRef Id="SocketLua" />
+      <ComponentRef Id="MimeLua" />
+      <ComponentRef Id="Ltn12Lua" />
+      <ComponentRef Id="SocketCoreDll" />
+      <ComponentRef Id="SocketFtp" />
+      <ComponentRef Id="SocketHeaders" />
+      <ComponentRef Id="SocketHttp" />
+      <ComponentRef Id="SocketTp" />
+      <ComponentRef Id="SocketUrl" />
+      <ComponentRef Id="MimeCoreDll" />
+    </Feature>
+
+  </Product>
+</Wix>

windows/pkg/install_service.bat

@@ -0,0 +1,24 @@
+@echo off
+REM Install lmcp as a Windows service using NSSM (Non-Sucking Service Manager)
+REM Download nssm from https://nssm.cc if not present
+
+if not exist "%~dp0nssm.exe" (
+    echo ERROR: nssm.exe not found in %~dp0
+    echo Download from https://nssm.cc and place nssm.exe here.
+    exit /b 1
+)
+
+set INSTALL_DIR=%~dp0
+set SERVICE_NAME=lmcp
+
+echo Installing lmcp as Windows service...
+%INSTALL_DIR%nssm.exe install %SERVICE_NAME% "%INSTALL_DIR%lua\lua.exe" "%INSTALL_DIR%server.lua"
+%INSTALL_DIR%nssm.exe set %SERVICE_NAME% AppDirectory "%INSTALL_DIR%"
+%INSTALL_DIR%nssm.exe set %SERVICE_NAME% AppEnvironmentExtra "LMCP_PORT=8080"
+%INSTALL_DIR%nssm.exe set %SERVICE_NAME% DisplayName "lmcp MCP Server"
+%INSTALL_DIR%nssm.exe set %SERVICE_NAME% Description "Lightweight MCP server in Lua"
+%INSTALL_DIR%nssm.exe set %SERVICE_NAME% Start SERVICE_AUTO_START
+%INSTALL_DIR%nssm.exe start %SERVICE_NAME%
+
+echo Done. Service '%SERVICE_NAME%' installed and started.
+echo Check: sc query %SERVICE_NAME%

windows/pkg/start.bat

@@ -0,0 +1,7 @@
+@echo off
+REM lmcp — Lua MCP Server
+REM Start the server on port 8080 (or LMCP_PORT if set)
+cd /d "%~dp0"
+if not defined LMCP_PORT set LMCP_PORT=8080
+echo Starting lmcp on port %LMCP_PORT%...
+lua\lua.exe server.lua

windows/sync.sh

@@ -0,0 +1,25 @@
+#!/bin/sh
+# windows/sync.sh — refresh windows/pkg/ from root .lua sources (issue #18).
+#
+# Run BEFORE invoking the WiX build so the MSI bundles whatever is in
+# master. The .lua files in windows/pkg/ are regenerated on every run
+# and are gitignored — never edit them directly.
+#
+# Idempotent: re-running just re-copies. Safe to call from a Makefile,
+# a CI step, or by hand before `candle.exe + light.exe`.
+
+set -eu
+
+here=$(dirname "$(readlink -f "$0")")
+root=$(cd "$here/.." && pwd)
+
+for f in lmcp.lua server.lua json.lua; do
+    if [ ! -f "$root/$f" ]; then
+        echo "windows/sync.sh: missing source $root/$f" >&2
+        exit 1
+    fi
+    cp "$root/$f" "$here/pkg/$f"
+    echo "  synced $f"
+done
+
+echo "windows/sync.sh: done — pkg/ matches root .lua at $(date +%Y-%m-%dT%H:%M:%S)"