feat: implement history split functionality to optimize context usage and add corresponding UI settings

webui/src/locales/en.json

@@ -379,6 +379,12 @@
         "behaviorTitle": "Behavior",
         "responsesTTL": "Responses store TTL (seconds)",
         "embeddingsProvider": "Embeddings provider",
+        "historySplitTitle": "History Split",
+        "historySplitDesc": "Pack earlier turns into an attached HISTORY.txt so the model reads the file first and then continues from the latest user request.",
+        "historySplitEnabled": "Enable history split",
+        "historySplitEnabledDesc": "Enabled by default. Turning this off falls back to normal full-context requests.",
+        "historySplitTriggerAfterTurns": "Trigger threshold (user turns)",
+        "historySplitTriggerHelp": "Default is 1, which means history split starts from the second turn.",
         "compatibilityTitle": "Compatibility",
         "compatibilityDesc": "Compatibility controls that keep stream output closer to the wire format or safer for the web UI.",
         "stripReferenceMarkers": "Strip [reference:N] markers",

webui/src/locales/zh.json

@@ -379,6 +379,12 @@
         "behaviorTitle": "行为设置",
         "responsesTTL": "Responses 缓存 TTL（秒）",
         "embeddingsProvider": "Embeddings Provider",
+        "historySplitTitle": "历史拆分",
+        "historySplitDesc": "将更早的对话整理成 HISTORY.txt 上传，让模型优先读取历史文件，再结合最新一轮继续回答。",
+        "historySplitEnabled": "启用历史拆分",
+        "historySplitEnabledDesc": "默认开启。关闭后会恢复为普通的完整上下文提交。",
+        "historySplitTriggerAfterTurns": "触发阈值（用户回合数）",
+        "historySplitTriggerHelp": "默认值为 1，表示从第二轮开始拆分历史。",
         "compatibilityTitle": "兼容性设置",
         "compatibilityDesc": "用于控制输出格式兼容性，避免把模型原始流里的标记直接暴露到前端。",
         "stripReferenceMarkers": "移除 [reference:N] 标记",