release: version 1.0.6

### Features
- Add NanobotServing - lightweight nanobot SDK integration
- CLI request retry mechanism with tqdm progress display
- Configurable cache_max_size_gb for S3DataSource/S3Storage

### Refactor
- LRUCacheManager cache architecture refactoring
- Unified Serving class naming with agent health check
- Rename temp_dir to cache_dir for consistency

### Fix
- Cache size configuration issues
- langkit dependency conflict
- Type annotations

### Docs
- Add usage examples for NanobotServing, CLI retry, and cache config
- Update README.md and README-zh.md

Author: User

CHANGELOG.md

@@ -7,6 +7,46 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ---
 
+## [1.0.6] - 2026-04-08
+
+### Added
+
+- **NanobotServing** (`5fe27e4`)
+  - 新增 `NanobotServing` - 基于 nanobot Python SDK 的轻量级 Serving 类
+  - 新增 `CLINanobotServing` 测试及 API 配置支持
+
+- **CLI 请求增强** (`4502473`)
+  - 添加 CLI 请求重试机制
+  - 添加 tqdm 进度显示
+
+- **缓存配置优化** (`e696b26`, `524e002`)
+  - `S3DataSource` 和 `S3Storage` 的 `cache_max_size_gb` 改为可配置参数
+
+### Changed
+
+- **缓存架构重构** (`ac516d2`)
+  - 重构 `LRUCacheManager` 缓存架构
+  - 将缓存逻辑从 `DataParser` 层移至 `Storage/DataSource` 层
+
+- **命名统一** (`55a1b50`, `a8d4fe4`)
+  - `S3Storage` 的 `temp_dir` 参数改为 `cache_dir`，统一命名
+  - `nanobot_serving` 重命名为 `cli_nanobot_serving` 以对齐命名规范
+  - 重构 `Serving` 类命名并添加 agent 健康检查
+
+### Fixed
+
+- **缓存大小配置** (`722f1a5`, `66b96e4`)
+  - 修复 cache size 配置问题
+  - 修复类型注解问题
+
+- **依赖冲突** (`ccfb7af`)
+  - 解决 langkit 依赖冲突问题
+
+- **临时目录支持** (`40ca1ae`)
+  - 为 `DataParser` 添加自定义临时目录支持
+
+---
+
 ## [1.0.5] - 2026-04-07
 
 ### Fixed

README-zh.md

@@ -629,6 +629,70 @@ serving = create_openclaw_serving(
 responses = serving.generate_from_input(["问题 1", "问题 2"])
 ```
 
+### NanobotServing - 轻量级 Nanobot SDK 集成
+
+```python
+from dataflow.serving import NanobotServing
+
+# 创建 NanobotServing
+serving = NanobotServing(
+    model_name="nanobot-model",
+    max_workers=4,
+    # 可选：API 配置
+    api_key="your-api-key",
+    base_url="https://api.nanobot.example.com",
+)
+
+# 生成响应
+responses = serving.generate_from_input(
+    ["Prompt 1", "Prompt 2"],
+    temperature=0.7,
+    max_tokens=1024,
+)
+```
+
+### CLI 请求重试和进度显示
+
+```python
+from dataflow.serving import CLIOpenClawServing
+
+serving = CLIOpenClawServing(
+    agent_id="main",
+    timeout=600,
+    max_workers=4,
+    # 重试配置
+    retry_times=3,
+    retry_delay=5,  # 秒
+    # 进度显示
+    show_progress=True,
+)
+
+responses = serving.generate_from_input(["问题 1", "问题 2"])
+```
+
+### 缓存配置
+
+```python
+from dataflow.utils.storage import S3DataSource, S3Storage
+
+# 配置缓存大小（默认：10GB）
+data_source = S3DataSource(
+    endpoint="https://s3.example.com",
+    ak="xxx",
+    sk="xxx",
+    s3_paths=["s3://bucket/data/"],
+    format_type="jsonl",
+    cache_dir="/data/cache",
+    cache_max_size_gb=50.0,  # 50GB 缓存
+)
+
+storage = S3Storage(
+    data_source=data_source,
+    id_key="id",
+    cache_max_size_gb=50.0,  # 与 DataSource 保持一致
+)
+```
+
 ---
 
 ## 7. 新增算子使用教程 (v1.0.3)

README.md

@@ -629,7 +629,7 @@ pipeline.compile()
 pipeline.forward(max_parallelism=4)  # Concurrent execution
 ```
 
-### Serving Usage Example
+### 2.2 Serving Usage Example
 
 ```python
 # OpenClaw CLI integration
@@ -643,6 +643,70 @@ serving = create_openclaw_serving(
 responses = serving.generate_from_input(["Question 1", "Question 2"])
 ```
 
+### 2.3 NanobotServing - Lightweight Nanobot SDK Integration
+
+```python
+from dataflow.serving import NanobotServing
+
+# Create NanobotServing
+serving = NanobotServing(
+    model_name="nanobot-model",
+    max_workers=4,
+    # Optional: API configuration
+    api_key="your-api-key",
+    base_url="https://api.nanobot.example.com",
+)
+
+# Generate responses
+responses = serving.generate_from_input(
+    ["Prompt 1", "Prompt 2"],
+    temperature=0.7,
+    max_tokens=1024,
+)
+```
+
+### 2.4 CLI Request with Retry and Progress
+
+```python
+from dataflow.serving import CLIOpenClawServing
+
+serving = CLIOpenClawServing(
+    agent_id="main",
+    timeout=600,
+    max_workers=4,
+    # Retry configuration
+    retry_times=3,
+    retry_delay=5,  # seconds
+    # Progress display
+    show_progress=True,
+)
+
+responses = serving.generate_from_input(["Question 1", "Question 2"])
+```
+
+### 2.5 Cache Configuration
+
+```python
+from dataflow.utils.storage import S3DataSource, S3Storage
+
+# Configure cache size (default: 10GB)
+data_source = S3DataSource(
+    endpoint="https://s3.example.com",
+    ak="xxx",
+    sk="xxx",
+    s3_paths=["s3://bucket/data/"],
+    format_type="jsonl",
+    cache_dir="/data/cache",
+    cache_max_size_gb=50.0,  # 50GB cache
+)
+
+storage = S3Storage(
+    data_source=data_source,
+    id_key="id",
+    cache_max_size_gb=50.0,  # Same as DataSource
+)
+```
+
 ---
 
 ## 7. New Operators Tutorial (v1.0.3)

dataflow/version.py

@@ -1,4 +1,4 @@
-__version__ = '1.0.5'
+__version__ = '1.0.6'
 short_version = __version__
 
 def parse_version_info(version_str):