diff --git a/src/meshcore_cli/history_store.py b/src/meshcore_cli/history_store.py
index 8b13789..2ea8d5f 100644
--- a/src/meshcore_cli/history_store.py
+++ b/src/meshcore_cli/history_store.py
@@ -1 +1,68 @@
 
+# -*- coding: utf-8 -*-
+"""
+HistoryStore: jednoduché perzistentní úložiště zpráv pro meshcore-cli
+- ukládá do ~/.config/meshcore/history/<device_name>/messages.jsonl
+- poskytuje metody append(record) a tail(n)
+"""
+import json
+import os
+from datetime import datetime, timezone
+from typing import List, Dict
+
+
+class HistoryStore:
+    def __init__(self, node_name: str, base_dir: str = None, rotate_daily: bool = False):
+        self.node_name = node_name or "unknown"
+        self.base_dir = base_dir or os.path.expanduser(f"~/.config/meshcore/history/{self.node_name}")
+        self.rotate_daily = rotate_daily
+        os.makedirs(self.base_dir, exist_ok=True)
+
+    def _file_path(self) -> str:
+        if self.rotate_daily:
+            day = datetime.now(timezone.utc).strftime("%Y-%m-%d")
+            return os.path.join(self.base_dir, f"messages-{day}.jsonl")
+        return os.path.join(self.base_dir, "messages.jsonl")
+
+    def append(self, record: Dict) -> None:
+        """
+        Přidá záznam do historie; automaticky doplní timestamp (UTC) a node_name.
+        """
+        rec = dict(record)
+        rec.setdefault("ts", datetime.now(timezone.utc).isoformat())
+        rec.setdefault("node_name", self.node_name)
+        fp = self._file_path()
+        with open(fp, "a", encoding="utf-8") as f:
+            f.write(json.dumps(rec, ensure_ascii=False) + "\n")
+
+    def tail(self, n: int = 50) -> List[Dict]:
+        """
+        Vrátí posledních n záznamů z historie.
+        """
+        fp = self._file_path()
+        if not os.path.exists(fp):
+            return []
+        # jednoduchý "tail": čte bloky od konce; pro běžné velikosti souboru dostačující
+        with open(fp, "rb") as f:
+            f.seek(0, os.SEEK_END)
+            size = f.tell()
+            block = 2048
+            data = b""
+            lines = []
+            while len(lines) <= n and size > 0:
+                size = max(0, size - block)
+                f.seek(size)
+                data = f.read(block) + data
+                lines = data.splitlines()
+            tail = lines[-n:] if len(lines) >= n else lines
+            out = []
+            for l in tail:
+                if not l.strip():
+                    continue
+                try:
+                    out.append(json.loads(l.decode("utf-8")))
+                except Exception:
+                    # poškozená řádka – přeskočit
+                    pass
+            return out
+