feat: Wave 5 — sysinfo speaks, smart reminders, outlook COM, time tracker, WOL, conversation tools

Voice sysinfo overhaul (6 scripts: battery/cpu/ram/disk/time/all):
- Now SPEAK the result via jarvis.cmd.ok instead of notify-only. Toast
  still fires for visual reference, but real Jarvis tells you out loud.
- `all` joins lines with commas so TTS doesn't over-pause on linebreaks.

Smart reminders (`reminders/set.lua` rewrite):
- Was: detached `Start-Sleep` PowerShell subprocesses that died on
  jarvis-app restart and spawned a hidden process per timer.
- Now: routes through `jarvis.scheduler.add(... type=speak ...)`. Persists
  to schedule.json, survives restart, no zombie processes. Trade-off:
  sub-minute timers round up to 1 min (rare in voice UX).
- Drops UA triggers from earlier UA-removal sweep.

Translation polish (`translate/translate.lua` + `clipboard.lua`):
- Drops UA target language; adds Italian, Polish, Turkish, French ru
  triggers + English target detectors.
- `language_iso()` helper picks the SAPI voice locale for the TARGET
  language (German translation → German SAPI voice if installed).

LLM system prompt (config.rs):
- New LLM_SYSTEM_PROMPT_EN counterpart. Both languages now mention the
  assistant's tooling (memory / scheduler / profiles / macros / vision /
  clipboard) so the LLM knows it can DO things, not suggest the user
  install something.
- `get_llm_system_prompt(lang)` picks EN/RU.

Conversation pack (NEW `conversation/`):
- 2 commands. `conversation.summary` ("о чём мы говорили") pulls the
  last 12 turns from persisted history, asks LLM for a 1-3 sentence
  recap in Jarvis tone. Falls back gracefully when offline.
- `conversation.repeat` ("повтори") re-speaks the last assistant message.
- New Lua API `jarvis.llm_history()` returns array of {role, content}
  tables, excluding the system prompt.

WOL pack (NEW `wol/`):
- Voice "разбуди сервер" / "wake server" fires a magic packet at a MAC
  stored in long-term memory under `wol_<alias>` (default alias "server").
- Setup via voice: "запомни wol_server AA:BB:CC:DD:EE:FF". Falls back to
  `JARVIS_WOL_TARGET` env if memory empty. Handles every standard MAC
  format (colons / dashes / dots / bare). Broadcast on UDP 9 via
  PowerShell System.Net.Sockets.UdpClient.

Outlook COM pack (NEW `outlook/`, via subagent):
- 4 commands: unread_count, latest, send_clipboard, summarize_inbox.
- PowerShell COM bridge `_outlook.ps1` with tab-delimited line protocol.
- summarize_inbox calls LLM for a one-paragraph summary. Graceful
  failure when Outlook isn't running.

Time tracker pack (NEW `time_tracker/`, via subagent):
- 5 commands: start, stop, today, week, reset.
- Persists via `jarvis.state` (structured object: current_session_start
  + sessions array of {start, end}).
- Local-calendar "today" boundary, full Russian pluralisation
  (час/часа/часов, минута/минуты/минут, сессия/сессии/сессий).

Tests: 139 rust (no regression), 81 python (60 → 81, +14 time_tracker
+7 outlook). Release build green for jarvis-app and jarvis-cli.

Pack count: 88 → 92 rust packs.

resources/commands/conversation/command.toml

@@ -0,0 +1,45 @@
+# Conversation tools — pull the persisted LLM history for review or replay.
+
+[[commands]]
+id = "conversation.summary"
+type = "lua"
+script = "summary.lua"
+sandbox = "full"
+timeout = 20000
+
+[commands.phrases]
+ru = [
+    "о чём мы говорили",
+    "о чем мы говорили",
+    "напомни о чем мы говорили",
+    "напомни про что мы говорили",
+    "что мы недавно обсуждали",
+    "сделай выжимку разговора",
+]
+en = [
+    "what were we talking about",
+    "summarize our conversation",
+    "what did we discuss",
+    "recap our chat",
+]
+
+
+[[commands]]
+id = "conversation.repeat"
+type = "lua"
+script = "repeat_last.lua"
+sandbox = "minimal"
+timeout = 3000
+
+[commands.phrases]
+ru = [
+    "повтори",
+    "повтори последнее",
+    "что ты сказал",
+    "не расслышал",
+]
+en = [
+    "say that again",
+    "repeat",
+    "i didn't catch that",
+]

resources/commands/conversation/repeat_last.lua

@@ -0,0 +1,13 @@
+-- Repeat the last assistant message verbatim. Useful when the user didn't catch
+-- it (loud room, distraction, etc).
+
+local lang = jarvis.context.language
+local last = jarvis.llm_last_reply()
+
+if not last or last == "" then
+    return jarvis.cmd.not_found(lang == "ru"
+        and "Я ещё ничего не говорил, сэр."
+        or  "I haven't said anything yet, sir.")
+end
+
+return jarvis.cmd.ok(last)

resources/commands/conversation/summary.lua

@@ -0,0 +1,79 @@
+-- Summarise the last N turns of LLM conversation in 1-2 sentences.
+--
+-- Reads the persistent history via `jarvis.llm_history()`, builds a short
+-- recap prompt, asks the LLM. If there's no history yet or the LLM is
+-- unreachable, falls back to a polite "ничего не вспомнить" reply.
+
+local lang = jarvis.context.language
+
+local turns = jarvis.llm_history() or {}
+-- Take only the last 12 user/assistant turns to keep prompt small.
+local recent = {}
+local start = math.max(1, #turns - 11)
+for i = start, #turns do
+    if turns[i] then table.insert(recent, turns[i]) end
+end
+
+if #recent == 0 then
+    return jarvis.cmd.not_found(lang == "ru"
+        and "Мы пока ещё ни о чём не говорили, сэр."
+        or  "We haven't talked about anything yet, sir.")
+end
+
+-- Render history as a compact "Пользователь: ... / Я: ..." dialogue.
+local lines = {}
+for _, m in ipairs(recent) do
+    local who = m.role == "assistant" and (lang == "ru" and "Я" or "Me")
+                                       or  (lang == "ru" and "Пользователь" or "User")
+    table.insert(lines, who .. ": " .. (m.content or ""))
+end
+local transcript = table.concat(lines, "\n")
+
+local token = jarvis.system.env("GROQ_TOKEN")
+if not token or token == "" then
+    -- Offline path: just speak the user's last message back briefly.
+    return jarvis.cmd.ok(lang == "ru"
+        and ("Без сети, сэр. Последний раз вы спросили: " .. (recent[#recent].content or ""):sub(1, 200))
+        or  ("No network, sir. Last topic: " .. (recent[#recent].content or ""):sub(1, 200)))
+end
+
+local base  = jarvis.system.env("GROQ_BASE_URL")
+if not base or base == "" then base = "https://api.groq.com/openai/v1" end
+local model = jarvis.system.env("GROQ_MODEL")
+if not model or model == "" then model = "llama-3.3-70b-versatile" end
+
+local sys
+if lang == "ru" then
+    sys = "Ты — J.A.R.V.I.S. По ниже идущему диалогу с пользователем составь короткое "
+        .. "напоминание (1-3 предложения), о чём шла речь. Говори от первого лица в "
+        .. "стиле британского дворецкого. Не пересказывай дословно — сожми суть."
+else
+    sys = "You are J.A.R.V.I.S. From the dialogue below, write a brief recap "
+        .. "(1-3 sentences) of what we talked about. Use first person, British "
+        .. "butler tone. Don't quote verbatim — distil the essence."
+end
+
+local payload = {
+    model = model,
+    messages = {
+        { role = "system", content = sys },
+        { role = "user",   content = transcript },
+    },
+    max_tokens = 200,
+    temperature = 0.4,
+}
+
+local res = jarvis.http.post_json(base .. "/chat/completions", payload,
+                                  { Authorization = "Bearer " .. token })
+if not res.ok then
+    return jarvis.cmd.error(lang == "ru" and "LLM не отвечает." or "LLM didn't respond.")
+end
+
+local content = (res.body or ""):match('"content"%s*:%s*"(.-[^\\])"')
+if not content then
+    return jarvis.cmd.error(lang == "ru" and "Не смог разобрать ответ." or "Couldn't parse reply.")
+end
+content = content:gsub('\\n', ' '):gsub('\\"', '"'):gsub('\\\\', '\\')
+content = content:gsub("%s+", " "):gsub("^%s+", ""):gsub("%s+$", "")
+
+return jarvis.cmd.ok(content)