feat(agent): agent for article fact checking

README.md

@@ -541,6 +541,7 @@ Both patterns share the same configuration interface and are transparent to user
 **Built-in Agents:**
 - `AgentFactCheck`: LangChain-based fact-checking with autonomous search control
 - `AgentHallucination`: Custom workflow hallucination detection with adaptive context gathering
+- `ArticleFactChecker`: Two-phase article fact-checking — extracts verifiable claims then verifies each in parallel using web search and Arxiv, with configurable concurrency control
 
 **Quick Example:**
 
@@ -597,6 +598,7 @@ For detailed guidance on choosing and implementing agent patterns, see [Agent De
 - [Agent Development Guide](docs/agent_development_guide.md) - Comprehensive guide for creating custom agents and tools
 - [AgentHallucination Example](examples/agent/agent_hallucination_example.py) - Production agent example
 - [AgentFactCheck Example](examples/agent/agent_executor_example.py) - LangChain agent example
+- [ArticleFactChecker Example](examples/agent/agent_article_fact_checking_example.py) - Article-scale two-phase fact verification
 
 ## ⚙️ Execution Modes
 

README_zh-CN.md

@@ -534,6 +534,7 @@ Dingo 支持基于智能体的评估器，可以使用外部工具进行多步
 **内置智能体：**
 - `AgentFactCheck`: 基于 LangChain 的事实核查，自主搜索控制
 - `AgentHallucination`: 自定义工作流的幻觉检测，自适应上下文收集
+- `ArticleFactChecker`: 两阶段文章事实核查 —— 先提取可验证声明，再并发调用网络搜索与 Arxiv 逐条验证，支持可配置的并发控制
 
 **快速示例：**
 
@@ -590,6 +591,7 @@ class MyAgent(BaseAgent):
 - [智能体开发指南](docs/agent_development_guide.md)
 - [AgentHallucination 示例](examples/agent/agent_hallucination_example.py)
 - [AgentFactCheck LangChain示例](examples/agent/agent_executor_example.py)
+- [ArticleFactChecker 示例](examples/agent/agent_article_fact_checking_example.py) - 文章级两阶段事实核查
 
 ## 执行引擎
 

dingo/model/llm/agent/__init__.py

@@ -1,22 +1,24 @@
-"""
-Agent Framework for Dingo
-
-This package provides agent-based evaluation capabilities that extend LLMs with
-tool usage, multi-step reasoning, and adaptive context gathering.
-
-Key Components:
-- BaseAgent: Abstract base class for agent evaluators
-- Tool system: Registry and base classes for agent tools
-"""
-
-from dingo.model.llm.agent.base_agent import BaseAgent
-from dingo.model.llm.agent.tools import BaseTool, ToolConfig, ToolRegistry, get_tool, tool_register
-
-__all__ = [
-    'BaseAgent',
-    'BaseTool',
-    'ToolConfig',
-    'ToolRegistry',
-    'get_tool',
-    'tool_register',
-]
+"""
+Agent Framework for Dingo
+
+This package provides agent-based evaluation capabilities that extend LLMs with
+tool usage, multi-step reasoning, and adaptive context gathering.
+
+Key Components:
+- BaseAgent: Abstract base class for agent evaluators
+- Tool system: Registry and base classes for agent tools
+"""
+
+from dingo.model.llm.agent.agent_article_fact_checker import ArticleFactChecker
+from dingo.model.llm.agent.base_agent import BaseAgent
+from dingo.model.llm.agent.tools import BaseTool, ToolConfig, ToolRegistry, get_tool, tool_register
+
+__all__ = [
+    'ArticleFactChecker',
+    'BaseAgent',
+    'BaseTool',
+    'ToolConfig',
+    'ToolRegistry',
+    'get_tool',
+    'tool_register',
+]