lite-agent 0.9.0__tar.gz → 0.11.0__tar.gz

{lite_agent-0.9.0 → lite_agent-0.11.0}/.claude/settings.local.json

@@ -18,7 +18,10 @@
       "Bash(uv run:*)",
       "Bash(mkdir:*)",
       "Bash(sed:*)",
-      "Bash(mv:*)"
+      "Bash(mv:*)",
+      "Bash(find:*)",
+      "WebSearch",
+      "Bash(uv list:*)"
     ],
     "deny": []
   }

{lite_agent-0.9.0 → lite_agent-0.11.0}/CHANGELOG.md

@@ -1,3 +1,37 @@
+## v0.11.0
+
+[v0.10.0...v0.11.0](https://github.com/Jannchie/lite-agent/compare/v0.10.0...v0.11.0)
+
+### :sparkles: Features
+
+- **client**: add typed dicts for reasoning config and refactor parsing logic - By [Jannchie](mailto:jannchie@gmail.com) in [8d125b4](https://github.com/Jannchie/lite-agent/commit/8d125b4)
+
+### :adhesive_bandage: Fixes
+
+- **message-builder**: ensure correct meta types and safer field access - By [Jannchie](mailto:jannchie@gmail.com) in [6c29c43](https://github.com/Jannchie/lite-agent/commit/6c29c43)
+
+## v0.10.0
+
+[v0.9.0...v0.10.0](https://github.com/Jannchie/lite-agent/compare/v0.9.0...v0.10.0)
+
+### :rocket: Breaking Changes
+
+- **reasoning-config**: streamline client reasoning config and improve error handling - By [Jannchie](mailto:jannchie@gmail.com) in [23c5d01](https://github.com/Jannchie/lite-agent/commit/23c5d01)
+- **runner**: remove deprecated run_continue methods - By [Jannchie](mailto:jannchie@gmail.com) in [37a8675](https://github.com/Jannchie/lite-agent/commit/37a8675)
+
+### :adhesive_bandage: Fixes
+
+- **chat-display**: separate label and content rows in messages - By [Jannchie](mailto:jannchie@gmail.com) in [06a6405](https://github.com/Jannchie/lite-agent/commit/06a6405)
+- **runner**: update meta fields handling in update_meta - By [Jannchie](mailto:jannchie@gmail.com) in [254f610](https://github.com/Jannchie/lite-agent/commit/254f610)
+
+### :art: Refactors
+
+- **messages**: modularize message handling and conversion - By [Jannchie](mailto:jannchie@gmail.com) in [acad106](https://github.com/Jannchie/lite-agent/commit/acad106)
+
+### :memo: Documentation
+
+- **docs**: expand examples and clarify message types - By [Jannchie](mailto:jannchie@gmail.com) in [9f3a5d7](https://github.com/Jannchie/lite-agent/commit/9f3a5d7)
+
 ## v0.9.0
 
 [v0.8.0...v0.9.0](https://github.com/Jannchie/lite-agent/compare/v0.8.0...v0.9.0)

{lite_agent-0.9.0 → lite_agent-0.11.0}/CLAUDE.md

@@ -38,17 +38,20 @@ uv run <command>         # Run command in project environment
 ### Running Examples
 
 ```bash
-uv run python examples/basic.py
-uv run python examples/handoffs.py  
-uv run python examples/chat_display_demo.py
-uv run python examples/context.py
-uv run python examples/terminal.py
-uv run python examples/translate/main.py
+uv run python examples/basic.py                    # Simple agent with tool calling
+uv run python examples/handoffs.py                 # Agent-to-agent transfers
+uv run python examples/chat_display_demo.py        # Rich console output
+uv run python examples/context.py                  # Context passing to tools
+uv run python examples/terminal.py                 # Terminal-based interaction
+uv run python examples/translate/main.py           # Translation agent example
+uv run python examples/streaming_demo.py           # Streaming responses demo
+uv run python examples/response_api_example.py     # Response API format demo
+uv run python scripts/record_chat_messages.py      # Record conversations for testing
 ```
 
 ## Project Architecture
 
-LiteAgent is a lightweight AI agent framework built on top of LiteLLM. The core architecture consists of:
+LiteAgent is a lightweight AI agent framework designed for flexibility with any LLM provider. The core architecture consists of:
 
 ### Core Components
 
@@ -127,6 +130,16 @@ The framework supports two OpenAI API modes:
 
 Set via `Runner(agent, api="completion")` or `Runner(agent, api="responses")`.
 
+### Message Types and Streaming
+
+The framework provides rich message types supporting both text and structured content:
+
+- **Text messages**: Simple string content for basic interactions
+- **Tool calls**: Structured function calls with parameters and results
+- **Agent transfers**: Built-in support for handoffs between specialized agents
+- **Rich content**: Support for complex message structures via Pydantic models
+- **Streaming chunks**: Real-time processing of LLM responses with granular event types
+
 ### Development Notes
 
 - Project uses strict ruff linting with `select = ["ALL"]` and specific ignores
@@ -135,7 +148,8 @@ Set via `Runner(agent, api="completion")` or `Runner(agent, api="responses")`.
 - Mock conversations stored in `tests/mocks/` as JSONL files for reproducible testing
 - Examples in `examples/` directory demonstrate various usage patterns
 - Template system uses Jinja2 for dynamic instruction generation (`src/lite_agent/templates/`)
-- Requires `litellm` as core dependency for LLM interactions
+- Does NOT directly depend on `litellm` - works with any compatible LLM client via `BaseLLMClient` interface
 - Chat display functionality uses `rich` library for formatted console output
+- Uses `pyright` for type checking with custom configuration excluding examples and temp directories
 
 The framework emphasizes simplicity and extensibility while maintaining full type safety and comprehensive streaming support.

{lite_agent-0.9.0 → lite_agent-0.11.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: lite-agent
-Version: 0.9.0
+Version: 0.11.0
 Summary: A lightweight, extensible framework for building AI agent.
 Author-email: Jianqi Pan <jannchie@gmail.com>
 License: MIT
@@ -13,6 +13,7 @@ Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
 Classifier: Topic :: Communications :: Chat
 Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Classifier: Topic :: Software Development :: Libraries :: Python Modules

{lite_agent-0.9.0 → lite_agent-0.11.0}/examples/basic.py

@@ -4,6 +4,8 @@ import logging
 from rich.logging import RichHandler
 
 from lite_agent.agent import Agent
+from lite_agent.chat_display import display_messages
+from lite_agent.client import LiteLLMClient
 from lite_agent.runner import Runner
 
 logging.basicConfig(
@@ -23,7 +25,7 @@ async def get_temperature(city: str) -> str:
 
 
 agent = Agent(
-    model="gpt-4.1-nano",
+    model=LiteLLMClient(model="gpt-5-nano", reasoning={"effort": "minimal"}),
     name="Weather Assistant",
     instructions="You are a helpful weather assistant. Before using tools, briefly explain what you are going to do. Provide friendly and informative responses.",
     tools=[get_temperature],
@@ -31,12 +33,12 @@ agent = Agent(
 
 
 async def main():
-    runner = Runner(agent, streaming=True, api="responses")
-    resp = await runner.run_until_complete(
+    runner = Runner(agent)
+    await runner.run_until_complete(
         "What is the temperature in New York?",
         includes=["usage", "assistant_message", "function_call", "function_call_output", "timing"],
     )
-    print(resp[0])
+    display_messages(runner.messages)
 
 
 if __name__ == "__main__":

{lite_agent-0.9.0 → lite_agent-0.11.0}/examples/knowledge/main.py

@@ -5,6 +5,7 @@ from rich.logging import RichHandler
 
 from lite_agent.agent import Agent
 from lite_agent.chat_display import display_messages
+from lite_agent.client import LiteLLMClient
 from lite_agent.runner import Runner
 
 logging.basicConfig(
@@ -45,18 +46,17 @@ def read_knowledge(knowledge: str) -> str:
 
 
 agent = Agent(
-    model="gpt-5-mini",
+    model=LiteLLMClient(model="gpt-5-nano", reasoning={"effort": "minimal"}),
     name="Assistant",
     instructions="你是一个有帮助的助手。",
     tools=[list_knowledges, read_knowledge],
-    reasoning={"effort": "minimal"},
 )
 
 
 async def main():
     runner = Runner(agent, streaming=True)
     await runner.run_until_complete(
-        "苹果的价格是多少？查询知识库告诉我答案。",
+        "苹果的价格是多少？仔细检查知识库告诉我答案。",
         includes=["usage", "assistant_message", "function_call", "function_call_output", "timing"],
     )
 

lite_agent-0.11.0/examples/reasoning_example.py

@@ -0,0 +1,127 @@
+import asyncio
+import logging
+
+from rich.logging import RichHandler
+
+from lite_agent.agent import Agent
+from lite_agent.client import LiteLLMClient
+
+logging.basicConfig(
+    level=logging.WARNING,
+    format="%(message)s",
+    datefmt="[%X]",
+    handlers=[RichHandler(rich_tracebacks=True)],
+)
+
+logger = logging.getLogger("lite_agent")
+logger.setLevel(logging.DEBUG)
+
+
+async def analyze_complex_problem(problem_description: str) -> str:
+    """Analyze a complex problem and return insights."""
+    return f"Analysis for: {problem_description}\n- Key factors identified\n- Potential solutions outlined\n- Risk assessment completed"
+
+
+async def demo_reasoning_configurations():
+    """演示不同的推理配置方法。"""
+    print("=== 推理配置演示 ===\n")
+
+    # 1. 使用 LiteLLMClient 设置推理强度（字符串形式）
+    print("1. 使用 LiteLLMClient 设置推理强度:")
+    agent_with_reasoning = Agent(
+        model=LiteLLMClient(model="gpt-4o-mini", reasoning="high"),
+        name="推理助手",
+        instructions="你是一个深度分析助手，使用仔细的推理来提供全面的分析。",
+    )
+    print(f"   客户端推理努力程度: {agent_with_reasoning.client.reasoning_effort}")
+    print(f"   客户端思考配置: {agent_with_reasoning.client.thinking_config}")
+
+    # 2. 使用 LiteLLMClient 进行更精细的控制（字典形式）
+    print("\n2. 使用 LiteLLMClient 进行精细控制:")
+    agent_with_thinking = Agent(
+        model=LiteLLMClient(
+            model="claude-3-5-sonnet-20241022",  # Anthropic模型支持thinking
+            reasoning={"type": "enabled", "budget_tokens": 2048},
+        ),
+        name="思考助手",
+        instructions="你是一个深思熟虑的助手。",
+    )
+    print(f"   客户端推理努力程度: {agent_with_thinking.client.reasoning_effort}")
+    print(f"   客户端思考配置: {agent_with_thinking.client.thinking_config}")
+
+    # 3. 使用 {"effort": "value"} 格式（推荐）
+    print('\n3. 使用 {"effort": "value"} 格式:')
+    agent_effort_reasoning = Agent(
+        model=LiteLLMClient(
+            model="o1-mini",  # OpenAI推理模型
+            reasoning={"effort": "medium"},
+        ),
+        name="努力推理助手",
+        instructions="你是一个高级推理助手。",
+    )
+    print(f"   客户端推理努力程度: {agent_effort_reasoning.client.reasoning_effort}")
+    print(f"   客户端思考配置: {agent_effort_reasoning.client.thinking_config}")
+
+    # 4. 使用布尔值设置推理（会默认使用medium级别）
+    print("\n4. 使用布尔值启用推理:")
+    agent_bool_reasoning = Agent(
+        model=LiteLLMClient(model="o1-mini", reasoning=True),  # 布尔值，会使用默认的medium级别
+        name="布尔推理助手",
+        instructions="你是一个高级推理助手。",
+    )
+    print(f"   客户端推理努力程度: {agent_bool_reasoning.client.reasoning_effort}")
+    print(f"   客户端思考配置: {agent_bool_reasoning.client.thinking_config}")
+
+    # 5. 演示运行时推理参数配置
+    print("\n5. 运行时推理参数配置:")
+    print("   - 推理配置现在只能在 LiteLLMClient 初始化时设置")
+    print("   - 如需动态调整，请创建不同的 Agent 实例")
+
+    # 注意：由于没有实际的API密钥，我们不运行真实的API调用
+    print("\n✓ 所有推理配置功能已成功设置！")
+
+
+async def main():
+    """主演示函数。"""
+    await demo_reasoning_configurations()
+
+    print("\n" + "=" * 60)
+    print("推理配置使用说明:")
+    print("=" * 60)
+    print("""
+1. reasoning 参数类型 (在 LiteLLMClient 中设置):
+   - 字符串: "minimal", "low", "medium", "high" -> reasoning_effort
+   - {"effort": "value"}: {"effort": "minimal"} -> reasoning_effort (推荐)
+   - 字典: {"type": "enabled", "budget_tokens": N} -> thinking_config
+   - 布尔: True -> "medium", False -> 不启用
+
+2. 使用方法:
+   ```python
+   # 方法1: 字符串形式
+   agent = Agent(
+       model=LiteLLMClient(model="gpt-4o-mini", reasoning="high")
+   )
+
+   # 方法2: {"effort": "value"} 形式（推荐）
+   agent = Agent(
+       model=LiteLLMClient(model="gpt-4o-mini", reasoning={"effort": "minimal"})
+   )
+
+   # 方法3: 字典形式 (Anthropic模型)
+   agent = Agent(
+       model=LiteLLMClient(
+           model="claude-3-5-sonnet-20241022",
+           reasoning={"type": "enabled", "budget_tokens": 2048}
+       )
+   )
+   ```
+
+3. 模型兼容性:
+   - OpenAI: o1, o3, gpt-4o-mini 等支持 reasoning_effort
+   - Anthropic: claude-3.5-sonnet 等支持 thinking
+   - 其他: 通过 LiteLLM 自动转换
+    """)
+
+
+if __name__ == "__main__":
+    asyncio.run(main())

{lite_agent-0.9.0 → lite_agent-0.11.0}/examples/stop_before_functions.py

@@ -60,7 +60,7 @@ async def main():
     print(f"\nHas require confirm tools: {await runner.has_require_confirm_tools()}")
 
     # Continue execution
-    resp = runner.run_continue_stream(
+    resp = runner.run(None,
         includes=["usage", "assistant_message", "function_call", "function_call_output"],
     )
     async for chunk in resp:
@@ -82,7 +82,7 @@ async def main():
     print(f"\nHas require confirm tools: {await runner2.has_require_confirm_tools()}")
 
     # Continue execution
-    resp = runner2.run_continue_stream(
+    resp = runner2.run(None,
         includes=["usage", "assistant_message", "function_call", "function_call_output"],
     )
     async for chunk in resp:
@@ -103,7 +103,7 @@ async def main():
     print(f"\nHas require confirm tools (decorator): {await runner3.has_require_confirm_tools()}")
 
     # Continue execution
-    resp = runner3.run_continue_stream(
+    resp = runner3.run(None,
         includes=["usage", "assistant_message", "function_call", "function_call_output"],
     )
     async for chunk in resp:

{lite_agent-0.9.0 → lite_agent-0.11.0}/examples/terminal.py

@@ -60,11 +60,11 @@ async def main():
                     validate_while_typing=False,
                 )
                 if user_input.lower() in {"y", "yes"}:
-                    response = runner.run_continue_stream()
+                    response = runner.run(None )
                     async for chunk in response:
                         await rich_channel.handle(chunk)
                 else:
-                    response = runner.run_continue_stream()
+                    response = runner.run(None )
             rich_channel.new_turn()
         except (EOFError, KeyboardInterrupt):
             break

{lite_agent-0.9.0 → lite_agent-0.11.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "lite-agent"
-version = "0.9.0"
+version = "0.11.0"
 description = "A lightweight, extensible framework for building AI agent."
 readme = "README.md"
 authors = [{ name = "Jianqi Pan", email = "jannchie@gmail.com" }]
@@ -29,6 +29,7 @@ classifiers = [
     "Programming Language :: Python :: 3.10",
     "Programming Language :: Python :: 3.11",
     "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13",
     "License :: OSI Approved :: MIT License",
     "Operating System :: OS Independent",
     "Intended Audience :: Developers",
@@ -90,10 +91,7 @@ ignore = [
     "ARG002",
     "S110",
 ]
-"examples/**/*" = [
-    "SLF001",
-    "S110",
-]
+"examples/**/*" = ["SLF001", "S110"]
 
 
 [tool.uv]

{lite_agent-0.9.0 → lite_agent-0.11.0}/scripts/record_chat_messages.py

@@ -42,7 +42,7 @@ async def main():
     resp = runner.run("What is the weather in New York? And what is the temperature there?")
     async for chunk in resp:
         logger.info(chunk)
-    resp = runner.run_continue_stream()
+    resp = runner.run(None)
     async for chunk in resp:
         logger.info(chunk)