lite-agent 0.2.0__tar.gz → 0.4.0__tar.gz

lite_agent-0.4.0/.claude/settings.local.json

@@ -0,0 +1,21 @@
+{
+  "$schema": "https://json.schemastore.org/claude-code-settings.json",
+  "permissions": {
+    "allow": [
+      "Bash(pytest:*)",
+      "mcp__context7__resolve-library-id",
+      "mcp__context7__get-library-docs",
+      "Bash(ruff check:*)",
+      "Bash(python:*)",
+      "Bash(ruff format:*)",
+      "Bash(rm:*)",
+      "Bash(grep:*)",
+      "mcp__ide__getDiagnostics",
+      "Bash(uv run pytest:*)",
+      "Bash(timeout:*)",
+      "Bash(pyright:*)",
+      "Bash(rg:*)"
+    ],
+    "deny": []
+  }
+}

{lite_agent-0.2.0 → lite_agent-0.4.0}/.vscode/launch.json

@@ -8,7 +8,7 @@
             "name": "Python main",
             "type": "debugpy",
             "request": "launch",
-            "program": "${workspaceFolder}/src/easy_agent/__main__.py",
+            "program": "${file}",
             "console": "integratedTerminal",
             "justMyCode": false
         }

lite_agent-0.4.0/CHANGELOG.md

@@ -0,0 +1,184 @@
+## v0.4.0
+
+[v0.3.0...v0.4.0](https://github.com/Jannchie/lite-agent/compare/v0.3.0...v0.4.0)
+
+### :rocket: Breaking Changes
+
+- **agent**: unify message handling using new messages format - By [Jannchie](mailto:jannchie@gmail.com) in [f31769a](https://github.com/Jannchie/lite-agent/commit/f31769a)
+- **chat-display**: remove display_chat_history usage and exports - By [Jannchie](mailto:jannchie@gmail.com) in [fca6ff7](https://github.com/Jannchie/lite-agent/commit/fca6ff7)
+- **chat-display**: rename rich_helpers to chat_display and update related imports - By [Jannchie](mailto:panjianqi@preferred.jp) in [8bbd83b](https://github.com/Jannchie/lite-agent/commit/8bbd83b)
+- **chunks**: rename tool_call and tool_call_result to function_call and function_call_output - By [Jannchie](mailto:jannchie@gmail.com) in [2821478](https://github.com/Jannchie/lite-agent/commit/2821478)
+- **function-call**: rename function_call_id to call_id - By [Jannchie](mailto:jannchie@gmail.com) in [d415fce](https://github.com/Jannchie/lite-agent/commit/d415fce)
+- **messages**: remove legacy function call messages and migrate to new assistant message structure - By [Jannchie](mailto:jannchie@gmail.com) in [c632629](https://github.com/Jannchie/lite-agent/commit/c632629)
+- **messages**: remove conversion helpers and fully switch to new message format - By [Jannchie](mailto:jannchie@gmail.com) in [75e47bf](https://github.com/Jannchie/lite-agent/commit/75e47bf)
+- **messages**: introduce new structured message types and unified message format - By [Jannchie](mailto:jannchie@gmail.com) in [cb8c091](https://github.com/Jannchie/lite-agent/commit/cb8c091)
+- **processors**: rename stream_chunk_processor to completion_event_processor - By [Jannchie](mailto:jannchie@gmail.com) in [80dedb9](https://github.com/Jannchie/lite-agent/commit/80dedb9)
+- **runner**: remove final_message type and streamline tool call handling - By [Jannchie](mailto:jannchie@gmail.com) in [ba91c29](https://github.com/Jannchie/lite-agent/commit/ba91c29)
+- **types**: rename chunk types to event types and update references - By [Jannchie](mailto:jannchie@gmail.com) in [d1354b3](https://github.com/Jannchie/lite-agent/commit/d1354b3)
+- **types**: remove finalmessagechunk type and usage - By [Jannchie](mailto:jannchie@gmail.com) in [6b3281e](https://github.com/Jannchie/lite-agent/commit/6b3281e)
+
+### :sparkles: Features
+
+- **chat-display**: add support for new message format - By [Jannchie](mailto:jannchie@gmail.com) in [a3fd1c6](https://github.com/Jannchie/lite-agent/commit/a3fd1c6)
+- **chat-display**: add meta stats, local time, and message performance info to chat display and message types - By [Jannchie](mailto:jannchie@gmail.com) in [bda3346](https://github.com/Jannchie/lite-agent/commit/bda3346)
+- **responses**: support new message format for response api - By [Jannchie](mailto:jannchie@gmail.com) in [e34a2ef](https://github.com/Jannchie/lite-agent/commit/e34a2ef)
+- **responses**: add response streaming support and client methods - By [Jannchie](mailto:jannchie@gmail.com) in [f484530](https://github.com/Jannchie/lite-agent/commit/f484530)
+- **rich-helpers**: add print_messages for compact output - By [Jannchie](mailto:panjianqi@preferred.jp) in [5a49203](https://github.com/Jannchie/lite-agent/commit/5a49203)
+- **runner**: support interactive tool call confirmation in terminal && improve chunk handling - By [Jannchie](mailto:jannchie@gmail.com) in [7164284](https://github.com/Jannchie/lite-agent/commit/7164284)
+
+### :adhesive_bandage: Fixes
+
+- **agent**: use isinstance for to_llm_dict checks - By [Jannchie](mailto:jannchie@gmail.com) in [e996785](https://github.com/Jannchie/lite-agent/commit/e996785)
+- **examples**: add content checks for dict messages - By [Jannchie](mailto:panjianqi@preferred.jp) in [94afcae](https://github.com/Jannchie/lite-agent/commit/94afcae)
+- **import**: update stream_handler imports && fix response api logic - By [Jannchie](mailto:jannchie@gmail.com) in [be192da](https://github.com/Jannchie/lite-agent/commit/be192da)
+- **input-image**: ensure file_id or image_url present and improve conversion handling - By [Jannchie](mailto:jannchie@gmail.com) in [d0bcdd5](https://github.com/Jannchie/lite-agent/commit/d0bcdd5)
+- **runner**: improve wait_for_user finish detection - By [Jannchie](mailto:jannchie@gmail.com) in [98435b0](https://github.com/Jannchie/lite-agent/commit/98435b0)
+- **runner**: fix final message processing to return responses - By [Jannchie](mailto:jannchie@gmail.com) in [e813f4a](https://github.com/Jannchie/lite-agent/commit/e813f4a)
+- **stream-chunk-processor**: construct usage event with correct usage type - By [Jannchie](mailto:jannchie@gmail.com) in [89b2cbe](https://github.com/Jannchie/lite-agent/commit/89b2cbe)
+- **tests**: improve type hints and patching mocks - By [Jannchie](mailto:jannchie@gmail.com) in [e4dfd64](https://github.com/Jannchie/lite-agent/commit/e4dfd64)
+- **tests**: update patch paths for litellm.acompletion - By [Jannchie](mailto:jannchie@gmail.com) in [897a284](https://github.com/Jannchie/lite-agent/commit/897a284)
+
+### :art: Refactors
+
+- **examples**: replace final_message with assistant_message in includes - By [Jannchie](mailto:jannchie@gmail.com) in [c5431ab](https://github.com/Jannchie/lite-agent/commit/c5431ab)
+- **runner**: simplify message appending logic && tidy code style - By [Jannchie](mailto:jannchie@gmail.com) in [412a8ec](https://github.com/Jannchie/lite-agent/commit/412a8ec)
+- **runner**: replace api handling with match statement && set default api to responses - By [Jannchie](mailto:jannchie@gmail.com) in [75c823a](https://github.com/Jannchie/lite-agent/commit/75c823a)
+- **stream-handlers**: rename litellm_stream_handler to litellm_completion_stream_handler && update imports and tests - By [Jannchie](mailto:jannchie@gmail.com) in [5b7ac92](https://github.com/Jannchie/lite-agent/commit/5b7ac92)
+- **tests**: rename tool_call types to function_call && remove unused usage tests - By [Jannchie](mailto:jannchie@gmail.com) in [1e0992f](https://github.com/Jannchie/lite-agent/commit/1e0992f)
+
+### :memo: Documentation
+
+- add claude guidance documentation - By [Jannchie](mailto:panjianqi@preferred.jp) in [a741494](https://github.com/Jannchie/lite-agent/commit/a741494)
+
+### :wrench: Chores
+
+- **dependencies**: update lock file - By [Jannchie](mailto:jannchie@gmail.com) in [042f6f3](https://github.com/Jannchie/lite-agent/commit/042f6f3)
+- **dev-deps**: add pytest-asyncio to dev dependencies - By [Jannchie](mailto:jannchie@gmail.com) in [57543cd](https://github.com/Jannchie/lite-agent/commit/57543cd)
+- **lint**: remove commented ruff exclude config - By [Jannchie](mailto:jannchie@gmail.com) in [25635dd](https://github.com/Jannchie/lite-agent/commit/25635dd)
+
+## v0.3.0
+
+[v0.2.0...v0.3.0](https://github.com/Jannchie/lite-agent/compare/v0.2.0...v0.3.0)
+
+### :rocket: Breaking Changes
+
+- **rich-helpers**: rename render_chat_history to print_chat_history and update imports - By [Jannchie](mailto:jannchie@gmail.com) in [3db6721](https://github.com/Jannchie/lite-agent/commit/3db6721)
+
+### :sparkles: Features
+
+- **agent**: use jinja2 templates for instruction messages - By [Jannchie](mailto:jannchie@gmail.com) in [f5ccbd5](https://github.com/Jannchie/lite-agent/commit/f5ccbd5)
+- **agent**: add completion_condition and task_done tool support - By [Jannchie](mailto:jannchie@gmail.com) in [28126b1](https://github.com/Jannchie/lite-agent/commit/28126b1)
+- **response-api**: add response api format for text and image input - By [Jannchie](mailto:jannchie@gmail.com) in [4b44f91](https://github.com/Jannchie/lite-agent/commit/4b44f91)
+- **rich-helpers**: add rich chat history renderer and chat summary utils - By [Jannchie](mailto:jannchie@gmail.com) in [5a570ce](https://github.com/Jannchie/lite-agent/commit/5a570ce)
+- **runner**: add set_chat_history to support full chat history replay and agent tracking - By [Jannchie](mailto:jannchie@gmail.com) in [fd2ecdc](https://github.com/Jannchie/lite-agent/commit/fd2ecdc)
+- **types**: enhance runner type system for flexible user_input - By [Jannchie](mailto:jannchie@gmail.com) in [3a2e9c5](https://github.com/Jannchie/lite-agent/commit/3a2e9c5)
+
+### :adhesive_bandage: Fixes
+
+- **examples**: check content key before modifying messages - By [Jannchie](mailto:jannchie@gmail.com) in [94b85a6](https://github.com/Jannchie/lite-agent/commit/94b85a6)
+
+### :art: Refactors
+
+- **agent**: abstract llm client and refactor model usage - By [Jannchie](mailto:jannchie@gmail.com) in [09c52c8](https://github.com/Jannchie/lite-agent/commit/09c52c8)
+- **client**: move llm client classes to separate module - By [Jannchie](mailto:jannchie@gmail.com) in [a70d163](https://github.com/Jannchie/lite-agent/commit/a70d163)
+- **rich-helpers-test**: rename render_chat_history to print_chat_history - By [Jannchie](mailto:jannchie@gmail.com) in [bf1ca8f](https://github.com/Jannchie/lite-agent/commit/bf1ca8f)
+
+### :lipstick: Styles
+
+- **agent**: inline error message formatting - By [Jannchie](mailto:jannchie@gmail.com) in [1453cc5](https://github.com/Jannchie/lite-agent/commit/1453cc5)
+- **tests**: add type ignore hints to assertions - By [Jannchie](mailto:jannchie@gmail.com) in [8d2a19d](https://github.com/Jannchie/lite-agent/commit/8d2a19d)
+
+### :memo: Documentation
+
+- **set-chat-history**: remove set_chat_history feature documentation - By [Jannchie](mailto:jannchie@gmail.com) in [2413351](https://github.com/Jannchie/lite-agent/commit/2413351)
+
+### :wrench: Chores
+
+- **examples**: rename test_consolidate_history to consolidate_history - By [Jannchie](mailto:jannchie@gmail.com) in [0f5b39a](https://github.com/Jannchie/lite-agent/commit/0f5b39a)
+- **tests**: remove legacy chat bubble and alignment tests - By [Jannchie](mailto:jannchie@gmail.com) in [46b23bc](https://github.com/Jannchie/lite-agent/commit/46b23bc)
+
+## v0.2.0
+
+[v0.1.0...v0.2.0](https://github.com/Jannchie/lite-agent/compare/v0.1.0...v0.2.0)
+
+### :rocket: Breaking Changes
+
+- **agent**: rename stream_async to completion - By [Jannchie](mailto:jannchie@gmail.com) in [dae25cc](https://github.com/Jannchie/lite-agent/commit/dae25cc)
+- **messages**: add support for function call message types and responses format conversion - By [Jannchie](mailto:jannchie@gmail.com) in [9e71ccc](https://github.com/Jannchie/lite-agent/commit/9e71ccc)
+- **project-structure**: rename package to lite-agent && update processor imports - By [Jannchie](mailto:jannchie@gmail.com) in [6e0747a](https://github.com/Jannchie/lite-agent/commit/6e0747a)
+- **runner**: rename run_stream to run throughout codebase - By [Jannchie](mailto:jannchie@gmail.com) in [525865f](https://github.com/Jannchie/lite-agent/commit/525865f)
+- **stream-handler**: rename litellm_raw to completion_raw - By [Jannchie](mailto:jannchie@gmail.com) in [9001f19](https://github.com/Jannchie/lite-agent/commit/9001f19)
+- **types**: remove AgentToolCallMessage usage and definition - By [Jannchie](mailto:jannchie@gmail.com) in [fb3592e](https://github.com/Jannchie/lite-agent/commit/fb3592e)
+- migrate chunk and message types to pydantic models && update runner and processors for model usage - By [Jannchie](mailto:jannchie@gmail.com) in [da0d6fc](https://github.com/Jannchie/lite-agent/commit/da0d6fc)
+
+### :sparkles: Features
+
+- **agent**: support dynamic handoff and parent transfer - By [Jannchie](mailto:jannchie@gmail.com) in [5019018](https://github.com/Jannchie/lite-agent/commit/5019018)
+- **agent**: add async tool call handling - By [Jannchie](mailto:jannchie@gmail.com) in [7de9103](https://github.com/Jannchie/lite-agent/commit/7de9103)
+- **context**: add context support for agent and runner - By [Jannchie](mailto:jannchie@gmail.com) in [e4885ee](https://github.com/Jannchie/lite-agent/commit/e4885ee)
+- **handoffs**: support multiple agents and add weather tools - By [Jannchie](mailto:jannchie@gmail.com) in [b9440ea](https://github.com/Jannchie/lite-agent/commit/b9440ea)
+- **handoffs**: add agent handoff and transfer functionality - By [Jannchie](mailto:jannchie@gmail.com) in [7655029](https://github.com/Jannchie/lite-agent/commit/7655029)
+- **message-transfer**: add message_transfer callback support to agent and runner && provide consolidate_history_transfer utility && add tests and usage examples - By [Jannchie](mailto:jannchie@gmail.com) in [16d4788](https://github.com/Jannchie/lite-agent/commit/16d4788)
+- **runner**: add record_to param to run_until_complete - By [Jannchie](mailto:jannchie@gmail.com) in [597bff6](https://github.com/Jannchie/lite-agent/commit/597bff6)
+- **runner**: support chat stream recording and improve sequence typing - By [Jannchie](mailto:jannchie@gmail.com) in [5c1d609](https://github.com/Jannchie/lite-agent/commit/5c1d609)
+- **runner**: add run_continue method and require confirm tool handling - By [Jannchie](mailto:jannchie@gmail.com) in [d130ebc](https://github.com/Jannchie/lite-agent/commit/d130ebc)
+- **testing**: add integration and unit tests for agent and file recording - By [Jannchie](mailto:jannchie@gmail.com) in [e5a58ce](https://github.com/Jannchie/lite-agent/commit/e5a58ce)
+- **tool**: add require_confirmation support for tool calls - By [Jannchie](mailto:panjianqi@preferred.jp) in [69625c8](https://github.com/Jannchie/lite-agent/commit/69625c8)
+- **tool-calls**: add require_confirm chunk type and flow - By [Jannchie](mailto:panjianqi@preferred.jp) in [b2d1654](https://github.com/Jannchie/lite-agent/commit/b2d1654)
+- **types**: add agent message and tool call types - By [Jannchie](mailto:jannchie@gmail.com) in [489e370](https://github.com/Jannchie/lite-agent/commit/489e370)
+
+### :adhesive_bandage: Fixes
+
+- **agent**: update tool target to completion - By [Jannchie](mailto:jannchie@gmail.com) in [6c8978c](https://github.com/Jannchie/lite-agent/commit/6c8978c)
+- **examples**: simplify temper agent instructions - By [Jannchie](mailto:jannchie@gmail.com) in [6632092](https://github.com/Jannchie/lite-agent/commit/6632092)
+- **loggers**: update logger name to lite_agent - By [Jannchie](mailto:jannchie@gmail.com) in [b4c59aa](https://github.com/Jannchie/lite-agent/commit/b4c59aa)
+- **types**: change tool_calls from sequence to list - By [Jannchie](mailto:jannchie@gmail.com) in [334c9c0](https://github.com/Jannchie/lite-agent/commit/334c9c0)
+
+### :art: Refactors
+
+- **agent**: rename prepare_messages to prepare_completion_messages - By [Jannchie](mailto:jannchie@gmail.com) in [02eeed6](https://github.com/Jannchie/lite-agent/commit/02eeed6)
+- **agent**: remove unused completions to responses converter - By [Jannchie](mailto:jannchie@gmail.com) in [f6432c6](https://github.com/Jannchie/lite-agent/commit/f6432c6)
+- **chunk-handler**: update type annotations and chunk types - By [Jannchie](mailto:panjianqi@preferred.jp) in [f2b95c5](https://github.com/Jannchie/lite-agent/commit/f2b95c5)
+- **chunk-handler**: extract helper functions for chunk processing - By [Jannchie](mailto:jannchie@gmail.com) in [083a164](https://github.com/Jannchie/lite-agent/commit/083a164)
+- **litellm**: extract chunk processing logic to function - By [Jannchie](mailto:jannchie@gmail.com) in [f31e459](https://github.com/Jannchie/lite-agent/commit/f31e459)
+- **litellm-stream**: decouple funcall from stream handler && adjust runner tool call logic - By [Jannchie](mailto:jannchie@gmail.com) in [2747e7e](https://github.com/Jannchie/lite-agent/commit/2747e7e)
+- **rich-channel**: extract rich channel to separate module && clean main - By [Jannchie](mailto:jannchie@gmail.com) in [a8be74d](https://github.com/Jannchie/lite-agent/commit/a8be74d)
+- **runner**: modularize runner methods and simplify message handling - By [Jannchie](mailto:jannchie@gmail.com) in [628a6b5](https://github.com/Jannchie/lite-agent/commit/628a6b5)
+- **tests**: parametrize tool call type tests and clean up redundant cases - By [Jannchie](mailto:jannchie@gmail.com) in [0791582](https://github.com/Jannchie/lite-agent/commit/0791582)
+- **types**: move chunk types to types.py && update type imports - By [Jannchie](mailto:panjianqi@preferred.jp) in [d2ebeec](https://github.com/Jannchie/lite-agent/commit/d2ebeec)
+
+### :lipstick: Styles
+
+- **tests**: improve formatting and consistency - By [Jannchie](mailto:jannchie@gmail.com) in [4342aa4](https://github.com/Jannchie/lite-agent/commit/4342aa4)
+
+### :memo: Documentation
+
+- add initial project readme - By [Jannchie](mailto:panjianqi@preferred.jp) in [f04ec7a](https://github.com/Jannchie/lite-agent/commit/f04ec7a)
+
+### :wrench: Chores
+
+- **ci**: switch to uv for dependency management and update to python 3.12 - By [Jannchie](mailto:panjianqi@preferred.jp) in [64d3085](https://github.com/Jannchie/lite-agent/commit/64d3085)
+- **ci**: add github actions workflow - By [Jannchie](mailto:panjianqi@preferred.jp) in [051be1b](https://github.com/Jannchie/lite-agent/commit/051be1b)
+- **deps**: update lock file - By [Jannchie](mailto:jannchie@gmail.com) in [9aa1e77](https://github.com/Jannchie/lite-agent/commit/9aa1e77)
+- **metadata**: rename project and reformat keywords array - By [Jannchie](mailto:jannchie@gmail.com) in [2659c4d](https://github.com/Jannchie/lite-agent/commit/2659c4d)
+- **metadata**: update ai topic classifier - By [Jannchie](mailto:jannchie@gmail.com) in [15d46ea](https://github.com/Jannchie/lite-agent/commit/15d46ea)
+- **rename**: rename project name - By [Jannchie](mailto:jannchie@gmail.com) in [84077a8](https://github.com/Jannchie/lite-agent/commit/84077a8)
+
+## v0.1.0
+
+[5859f296f4aa3bf9156e560e5bbd98b3d8795b0d...v0.1.0](https://github.com/Jannchie/lite-agent/compare/5859f296f4aa3bf9156e560e5bbd98b3d8795b0d...v0.1.0)
+
+### :sparkles: Features
+
+- **chunk-handler**: add content-delta and tool-call-delta chunk types && update includes list in runner - By [Jannchie](mailto:jannchie@gmail.com) in [a3ba7d1](https://github.com/Jannchie/lite-agent/commit/a3ba7d1)
+- **easy_agent**: add prompt toolkit dependency && update agent to use prompt session && improve chunk handling - By [Jannchie](mailto:jannchie@gmail.com) in [b408c14](https://github.com/Jannchie/lite-agent/commit/b408c14)
+
+### :art: Refactors
+
+- **main**: remove datetime imports and usage && add input validator - By [Jannchie](mailto:jannchie@gmail.com) in [eab45b3](https://github.com/Jannchie/lite-agent/commit/eab45b3)
+
+### :wrench: Chores
+
+- **import**: remove unused json tool import - By [Jannchie](mailto:jannchie@gmail.com) in [8de5975](https://github.com/Jannchie/lite-agent/commit/8de5975)
+- **lock**: update lock file - By [Jannchie](mailto:jannchie@gmail.com) in [64880d3](https://github.com/Jannchie/lite-agent/commit/64880d3)
+- **pyproject**: update version && improve description && expand keywords && add classifiers - By [Jannchie](mailto:jannchie@gmail.com) in [c3e8ee6](https://github.com/Jannchie/lite-agent/commit/c3e8ee6)

lite_agent-0.4.0/CLAUDE.md

@@ -0,0 +1,100 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## Development Commands
+
+### Testing
+
+```bash
+pytest                    # Run all tests
+pytest tests/unit/        # Run only unit tests
+pytest tests/integration/ # Run only integration tests
+pytest --cov             # Run with coverage
+```
+
+### Linting and Formatting
+
+```bash
+ruff check               # Run linter
+ruff format              # Format code
+```
+
+### Package Management
+
+```bash
+uv add lite-agent # Install from PyPI
+uv add --dev lite-agent # Install dev package
+```
+
+## Project Architecture
+
+LiteAgent is a lightweight AI agent framework built on top of LiteLLM. The core architecture consists of:
+
+### Core Components
+
+**Agent (`src/lite_agent/agent.py`)**
+
+- Central agent class that manages LLM interactions, tool calls, and message handling
+- Supports tool registration via `funcall` library for type-safe function calling
+- Handles agent handoffs (parent-child relationships) for complex workflows
+- Manages completion conditions ("stop" vs "call" for different termination behaviors)
+- Converts between OpenAI's Response API and Completion API message formats
+
+**Runner (`src/lite_agent/runner.py`)**
+
+- Orchestrates agent execution with streaming support
+- Manages conversation flow and message history
+- Handles tool call execution and agent transfers
+- Supports continuation from previous states and chat history management
+- Provides both streaming and batch execution modes
+
+**Type System (`src/lite_agent/types/`)**
+
+- Comprehensive Pydantic models for all message types and chunks
+- Supports both Response API and Completion API formats
+- Type-safe definitions for tool calls, chunks, and messages
+
+### Key Features
+
+**Tool Integration**
+
+- Uses `funcall` library for automatic tool schema generation from Python functions
+- Supports basic types, Pydantic models, and dataclasses as parameters
+- Dynamic tool registration for agent handoffs and control flow
+
+**Agent Handoffs**
+
+- Parent-child agent relationships for complex task delegation
+- Automatic `transfer_to_agent` and `transfer_to_parent` tool registration
+- Chat history tracking across agent transitions
+
+**Message Processing**
+
+- Bidirectional conversion between OpenAI API formats
+- Streaming chunk processing with configurable output filtering
+- Message transfer callbacks for preprocessing
+
+**Completion Modes**
+
+- `"stop"`: Traditional completion until model decides to stop
+- `"call"`: Completion until specific tool (`wait_for_user`) is called
+
+### Examples Directory Structure
+
+Examples demonstrate various usage patterns:
+
+- `basic.py`: Simple agent with tool calling
+- `handoffs.py`: Agent-to-agent transfers
+- `context.py`: Context passing to tools
+- `chat_display_demo.py`: Rich console output formatting
+- `channels/`: Channel-based communication patterns
+
+### Testing Architecture
+
+- **Unit tests**: Test individual components in isolation
+- **Integration tests**: Test full agent workflows with mocked LLM responses
+- **Performance tests**: Test memory usage and performance characteristics
+- **Mock system**: JSONL-based conversation recording/playback for deterministic testing
+
+The framework emphasizes simplicity and extensibility while maintaining full type safety and comprehensive streaming support.

{lite_agent-0.2.0 → lite_agent-0.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: lite-agent
-Version: 0.2.0
+Version: 0.4.0
 Summary: A lightweight, extensible framework for building AI agent.
 Author-email: Jianqi Pan <jannchie@gmail.com>
 License: MIT
@@ -18,7 +18,7 @@ Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Requires-Python: >=3.10
 Requires-Dist: aiofiles>=24.1.0
-Requires-Dist: funcall>=0.7.0
+Requires-Dist: funcall>=0.10.0
 Requires-Dist: prompt-toolkit>=3.0.51
 Requires-Dist: rich>=14.0.0
 Description-Content-Type: text/markdown

{lite_agent-0.2.0 → lite_agent-0.4.0}/examples/basic.py

@@ -4,6 +4,7 @@ import logging
 from rich.logging import RichHandler
 
 from lite_agent.agent import Agent
+from lite_agent.chat_display import display_messages
 from lite_agent.runner import Runner
 
 logging.basicConfig(
@@ -19,7 +20,6 @@ logger.setLevel(logging.DEBUG)
 
 async def get_temperature(city: str) -> str:
     """Get the temperature for a city."""
-    await asyncio.sleep(1)
     return f"The temperature in {city} is 25°C."
 
 
@@ -35,12 +35,13 @@ async def main():
     runner = Runner(agent)
     resp = runner.run(
         "What is the temperature in New York?",
-        includes=["final_message", "usage", "tool_call", "tool_call_result"],
-        record_to="tests/mocks/basic/1.jsonl",
+        includes=["usage", "assistant_message", "function_call", "function_call_output", "timing"],
     )
     async for chunk in resp:
         logger.info(chunk)
+    display_messages(runner.messages)
     print(runner.messages)
 
+
 if __name__ == "__main__":
     asyncio.run(main())

lite_agent-0.4.0/examples/basic_agent.py

@@ -0,0 +1,54 @@
+import asyncio
+import logging
+
+from rich.logging import RichHandler
+
+from lite_agent.agent import Agent
+
+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 get_temperature(city: str) -> str:
+    """Get the temperature for a city."""
+    await asyncio.sleep(1)
+    return f"The temperature in {city} is 25°C."
+
+
+async def get_whether(city: str) -> str:
+    """Get the weather for a city."""
+    await asyncio.sleep(1)
+    return f"The weather in {city} is sunny with a few clouds."
+
+
+agent = Agent(
+    model="gpt-4.1-nano",
+    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, get_whether],
+)
+
+
+async def main():
+    resp = await agent.completion(
+        [
+            {
+                "role": "user",
+                "content": "What is the temperature and whether in New York?",
+            },
+        ],
+    )
+    async for chunk in resp:
+        if chunk.type != "completion_raw":
+            logger.info(chunk)
+
+
+if __name__ == "__main__":
+    asyncio.run(main())

{lite_agent-0.2.0 → lite_agent-0.4.0}/examples/channels/rich_channel.py

@@ -1,31 +1,32 @@
 from rich.console import Console
 
-from lite_agent.types import AgentChunk, ContentDeltaChunk
+from lite_agent.types import AgentChunk, ContentDeltaEvent
 
 
 class RichChannel:
     def __init__(self) -> None:
         self.console = Console()
         self.map = {
-            "final_message": self.handle_final_message,
-            "tool_call": self.handle_tool_call,
-            "tool_call_result": self.handle_tool_call_result,
-            "tool_call_delta": self.handle_tool_call_delta,
+            "function_call": self.handle_tool_call,
+            "function_call_output": self.handle_tool_call_result,
+            "function_call_delta": self.handle_function_call_delta,
             "content_delta": self.handle_content_delta,
             "usage": self.handle_usage,
-            "require_confirm": self.handle_require_confirm,
         }
-        self.new_turn = True
+        self._new_turn = True
 
     async def handle(self, chunk: AgentChunk):
-        handler = self.map[chunk.type]
-        return await handler(chunk)
+        handler = self.map.get(chunk.type)
+        if handler is None:
+            return None
+        return await handler(chunk)  # type: ignore
 
-    async def handle_final_message(self, _chunk: AgentChunk):
+    def new_turn(self):
         print()
-        self.new_turn = True
+        self._new_turn = True
 
     async def handle_tool_call(self, chunk: AgentChunk):
+        print()
         name = getattr(chunk, "name", "<unknown>")
         arguments = getattr(chunk, "arguments", "")
         self.console.print(f"🛠️  [green]{name}[/green]([yellow]{arguments}[/yellow])")
@@ -35,16 +36,14 @@ class RichChannel:
         content = getattr(chunk, "content", "")
         self.console.print(f"🛠️  [green]{name}[/green] → [yellow]{content}[/yellow]")
 
-    async def handle_tool_call_delta(self, chunk: AgentChunk): ...
-    async def handle_content_delta(self, chunk: ContentDeltaChunk):
-        if self.new_turn:
+    async def handle_function_call_delta(self, chunk: AgentChunk): ...
+    async def handle_content_delta(self, chunk: ContentDeltaEvent):
+        if self._new_turn:
             self.console.print("🤖 ", end="")
-            self.new_turn = False
+            self._new_turn = False
         print(chunk.delta, end="", flush=True)
 
     async def handle_usage(self, chunk: AgentChunk):
         if False:
             usage = chunk.usage
             self.console.print(f"In: {usage.prompt_tokens}, Out: {usage.completion_tokens}, Total: {usage.total_tokens}")
-
-    async def handle_require_confirm(self, chunk: AgentChunk): ...

lite_agent-0.4.0/examples/chat_display_demo.py

@@ -0,0 +1,113 @@
+"""
+示例：使用 chat_display 美观显示聊天记录
+
+这个示例展示了如何使用 chat_display 模块中的函数来美观地显示聊天记录。
+"""
+
+import asyncio
+
+from lite_agent import display_messages
+from lite_agent.agent import Agent
+from lite_agent.runner import Runner
+
+
+def demo_tools():
+    """演示工具函数，用于测试函数调用显示。"""
+
+    def get_weather(city: str) -> str:
+        """获取指定城市的天气信息。"""
+        return f"The weather in {city} is sunny with 25°C"
+
+    def calculate(expression: str) -> str:
+        """计算数学表达式。"""
+        try:
+            result = eval(expression)  # noqa: S307
+        except Exception as e:
+            return f"Error calculating {expression}: {e}"
+        else:
+            return f"{expression} = {result}"
+
+    return [get_weather, calculate]
+
+
+async def create_sample_chat_history() -> Runner:
+    """创建一个包含各种消息类型的示例聊天历史。"""
+    # 创建 agent
+    agent = Agent(
+        model="gpt-4o-mini",
+        name="DemoAgent",
+        instructions="You are a helpful assistant that can provide weather information and perform calculations.",
+        tools=demo_tools(),
+    )
+
+    # 创建 runner
+    runner = Runner(agent=agent)
+
+    # 手动添加一些示例消息来展示不同的消息类型
+    runner.append_message({"role": "system", "content": "You are a helpful assistant."})
+    runner.append_message({"role": "user", "content": "Hello! Can you help me with some tasks?"})
+    runner.append_message({"role": "assistant", "content": "Of course! I'd be happy to help you. What would you like to do?"})
+    runner.append_message({"role": "user", "content": "What's the weather like in Tokyo?"})
+
+    # 添加函数调用消息
+    runner.append_message(
+        {
+            "type": "function_call",
+            "call_id": "call_123",
+            "name": "get_weather",
+            "arguments": '{"city": "Tokyo"}',
+            "content": "",
+        },
+    )
+
+    # 添加函数调用输出
+    runner.append_message(
+        {
+            "type": "function_call_output",
+            "call_id": "call_123",
+            "output": "The weather in Tokyo is sunny with 25°C",
+        },
+    )
+
+    runner.append_message({"role": "assistant", "content": "The weather in Tokyo is sunny with a temperature of 25°C. Is there anything else you'd like to know?"})
+    runner.append_message({"role": "user", "content": "Can you calculate 25 * 4 + 10?"})
+
+    # 添加另一个函数调用
+    runner.append_message(
+        {
+            "type": "function_call",
+            "call_id": "call_456",
+            "name": "calculate",
+            "arguments": '{"expression": "25 * 4 + 10"}',
+            "content": "",
+        },
+    )
+
+    runner.append_message(
+        {
+            "type": "function_call_output",
+            "call_id": "call_456",
+            "output": "25 * 4 + 10 = 110",
+        },
+    )
+
+    runner.append_message({"role": "assistant", "content": "The calculation 25 * 4 + 10 equals 110."})
+
+    return runner
+
+
+async def main():
+    """主函数：演示 chat_display 的使用。"""
+    print("🎨 Chat Display Demo\n")
+
+    # 创建示例聊天历史
+    runner = await create_sample_chat_history()
+
+    # print demo
+    display_messages(
+        runner.messages,
+    )
+
+
+if __name__ == "__main__":
+    asyncio.run(main())

{lite_agent-0.2.0 → lite_agent-0.4.0}/examples/confirm_and_continue.py

@@ -43,13 +43,13 @@ async def main():
     runner = Runner(agent)
     resp = runner.run(
         "What is the weather in New York? And what is the temperature there?",
-        includes=["final_message", "usage", "tool_call", "tool_call_result"],
+        includes=["usage", "assistant_message", "function_call", "function_call_output"],
         record_to="tests/mocks/confirm_and_continue/1.jsonl",
     )
     async for chunk in resp:
         logger.info(chunk)
     resp = runner.run_continue_stream(
-        includes=["final_message", "usage", "tool_call", "tool_call_result"],
+        includes=["usage", "assistant_message", "function_call", "function_call_output"],
         record_to="tests/mocks/confirm_and_continue/2.jsonl",
     )
     async for chunk in resp:

lite_agent-0.2.0/examples/test_consolidate_history.py → lite_agent-0.4.0/examples/consolidate_history.py

@@ -26,7 +26,7 @@ def test_consolidate_history():
     print("Consolidated messages count:", len(result))
     print("\nConsolidated content:")
     print("=" * 50)
-    if result and isinstance(result[0], dict):
+    if result and isinstance(result[0], dict) and "content" in result[0]:
         print(result[0]["content"])
     print("=" * 50)
 
@@ -38,7 +38,7 @@ def test_consolidate_history():
     single_message = [{"role": "user", "content": "Just a simple test"}]
     single_result = consolidate_history_transfer(single_message)
     print(f"\nSingle message result count: {len(single_result)}")
-    if single_result and isinstance(single_result[0], dict):
+    if single_result and isinstance(single_result[0], dict) and "content" in single_result[0]:
         print("Single message consolidated content:")
         print(single_result[0]["content"])