@oneciel-ai/claude-any 0.1.63 → 0.1.64

package/README.md

@@ -48,7 +48,7 @@ arguments through unchanged.
 
 Credits: One Ciel LLC
 
-Current version: `0.1.63`
+Current version: `0.1.64`
 
 ## Why This Exists
 
@@ -385,6 +385,15 @@ steps under that larger model's supervision.
 
 ## Changelog
 
+### 0.1.64
+
+- **Model-aware native auto-compact**: claude-any now injects
+  `CLAUDE_CODE_AUTO_COMPACT_WINDOW` at launch using the selected provider/model
+  context window, including the cached Ollama/Ollama Cloud model catalog. Smaller
+  custom models now let Claude Code's native auto-compact trigger against their
+  real context budget instead of falling back to Claude Code's generic 200K
+  assumption.
+
 ### 0.1.63
 
 - **Plan Mode stop guard**: when a non-Anthropic model is already in Plan Mode

package/claude_any.py

@@ -90,7 +90,7 @@ PROVIDER_LABELS = {
     "self-hosted-nim": "Self Hosted NIM",
 }
 APP_NAME = "Claude Any"
-VERSION = "0.1.63"
+VERSION = "0.1.64"
 CREDITS = "Credits: One Ciel LLC"
 
 LOG_LEVELS = {"SILENT": 0, "ERROR": 1, "WARN": 2, "INFO": 3, "DEBUG": 4, "TRACE": 5}
@@ -8774,6 +8774,13 @@ def claude_code_output_token_limit(provider: str, pcfg: dict[str, Any]) -> int |
     return None
 
 
+def claude_code_auto_compact_window(provider: str, pcfg: dict[str, Any]) -> int | None:
+    limit = context_limit_for_status(provider, pcfg)
+    if limit:
+        return limit
+    return None
+
+
 def claude_code_context_model_alias(provider: str, pcfg: dict[str, Any], model: str) -> str:
     model = strip_claude_context_suffix(model)
     limit = context_limit_for_status(provider, pcfg)
@@ -8790,6 +8797,9 @@ def apply_common_claude_env(provider: str, pcfg: dict[str, Any], env: dict[str,
     output_tokens = claude_code_output_token_limit(provider, pcfg)
     if output_tokens:
         env["CLAUDE_CODE_MAX_OUTPUT_TOKENS"] = str(output_tokens)
+    compact_window = claude_code_auto_compact_window(provider, pcfg)
+    if compact_window:
+        env["CLAUDE_CODE_AUTO_COMPACT_WINDOW"] = str(compact_window)
     advisor_model = str(pcfg.get("advisor_model") or "").strip()
     if advisor_model:
         env["CLAUDE_ANY_ADVISOR_MODEL"] = advisor_model
@@ -8880,6 +8890,7 @@ def cmd_env(_: argparse.Namespace) -> None:
         "CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS",
         "CLAUDE_CODE_ATTRIBUTION_HEADER",
         "CLAUDE_CODE_MAX_OUTPUT_TOKENS",
+        "CLAUDE_CODE_AUTO_COMPACT_WINDOW",
         "ANTHROPIC_MODEL",
         "ANTHROPIC_CUSTOM_MODEL_OPTION",
         "ANTHROPIC_DEFAULT_HAIKU_MODEL",

package/docs/README.ja.md

@@ -47,7 +47,7 @@ vLLM、NVIDIA hosted、self-hosted NIM を選択し、通常の Claude Code 引
 
 Credits: One Ciel LLC
 
-現在のバージョン: `0.1.63`
+現在のバージョン: `0.1.64`
 
 ## 作られた理由
 
@@ -351,6 +351,14 @@ Windows/Linux 管理、クリーンアップスクリプト、定期的なセキ
 
 ## 変更履歴
 
+### 0.1.64
+
+- **モデル context 対応の native auto-compact**: claude-any は起動時に、選択中の
+  provider/model の context window を使って `CLAUDE_CODE_AUTO_COMPACT_WINDOW`
+  を注入します。Ollama/Ollama Cloud ではディスクにキャッシュした model catalog
+  も利用するため、小さい custom model でも Claude Code の汎用 200K 仮定ではなく、
+  実際の context budget に合わせて native auto-compact が発火します。
+
 ### 0.1.63
 
 - **Plan Mode stop guard**: non-Anthropic モデルが Plan Mode 中に短い確認文だけで

package/docs/README.ko.md

@@ -47,7 +47,7 @@ NVIDIA hosted, self-hosted NIM을 선택하고, Claude Code의 일반 인자는
 
 Credits: One Ciel LLC
 
-현재 버전: `0.1.63`
+현재 버전: `0.1.64`
 
 ## 왜 만들었나
 
@@ -351,6 +351,14 @@ Windows 이벤트 로그 리뷰, 바이러스/랜섬웨어 침입 시도 정리,
 
 ## 변경 이력
 
+### 0.1.64
+
+- **모델 컨텍스트 인식 native auto-compact**: claude-any가 실행 시 선택된
+  provider/model의 context window를 기준으로 `CLAUDE_CODE_AUTO_COMPACT_WINDOW`를
+  주입합니다. Ollama/Ollama Cloud는 디스크에 캐시된 model catalog도 활용하므로,
+  작은 커스텀 모델도 Claude Code의 기본 200K 가정이 아니라 실제 context budget에
+  맞춰 native auto-compact가 발동됩니다.
+
 ### 0.1.63
 
 - **Plan Mode stop guard**: non-Anthropic 모델이 Plan Mode 안에서 짧은 확인

package/docs/README.zh.md

@@ -47,7 +47,7 @@ NIM，并把普通 Claude Code 参数原样传递。
 
 Credits: One Ciel LLC
 
-当前版本: `0.1.63`
+当前版本: `0.1.64`
 
 ## 为什么存在
 
@@ -337,6 +337,14 @@ Hermes 格式模型或部分较旧的 Qwen tool template。
 
 ## 更新日志
 
+### 0.1.64
+
+- **按模型上下文触发 native auto-compact**：claude-any 启动时会根据当前
+  provider/model 的 context window 注入 `CLAUDE_CODE_AUTO_COMPACT_WINDOW`。
+  Ollama/Ollama Cloud 会同时使用磁盘缓存的 model catalog，因此较小的 custom
+  model 也会按真实 context budget 触发 Claude Code 原生 auto-compact，而不是
+  退回到 Claude Code 通用的 200K 假设。
+
 ### 0.1.63
 
 - **Plan Mode stop guard**：当 non-Anthropic 模型已经处于 Plan Mode，却只输出

package/docs/manual.md

@@ -10,7 +10,7 @@ Code starts, while passing normal Claude Code arguments through unchanged.
 
 Credits: One Ciel LLC
 
-Current version: `0.1.63`
+Current version: `0.1.64`
 
 ## Install
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@oneciel-ai/claude-any",
-  "version": "0.1.63",
+  "version": "0.1.64",
   "description": "Claude Code provider selector for Anthropic, Ollama, Ollama Cloud, vLLM, NVIDIA hosted, and self-hosted NIM.",
   "license": "MIT",
   "author": "One Ciel LLC",