feat(memory): add reme long-term memory

[jinliyl]

feat(ad reme memory): implement ReMe-based long-term memory modules

• Add ReMePersonalMemory, ReMeTaskMemory, and ReMeToolMemory classes
• Implement base class ReMeBaseLongTermMemory with common functionality
• Add support for personal memory recording and retrieval with keywords
• Implement task memory for storing and retrieving execution trajectories
• Create tool memory for recording tool call results and generating guidelines
• Add proper error handling and logging with loguru- Include async context manager support for memory initialization- Add comprehensive docstrings and type hints
• Integrate with DashScope and OpenAI models for embedding and LLM APIs
• Provide example usage in personal memory demonstration script

[DavdGao]

Thanks for your contribution. Please see inline comments.

[Copilot]

Pull Request Overview

This PR adds ReMe (Reflection Memory) long-term memory support to AgentScope, introducing three specialized memory types: Personal, Task, and Tool memory. The implementation provides both tool function interfaces (for agent-driven memory management) and direct method interfaces (for programmatic control), with comprehensive test coverage and documentation.

Key changes:

• Added three ReMe memory implementations: ReMePersonalMemory, ReMeTaskMemory, and ReMeToolMemory
• Integrated ReMe memory types into AgentScope's memory system with dual interface design
• Added comprehensive test suite, examples, and documentation

Reviewed Changes

Copilot reviewed 15 out of 15 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
pyproject.toml	Added reme-ai dependency with Python 3.12+ constraint
src/agentscope/memory/__init__.py	Exported three new ReMe memory classes
src/agentscope/memory/reme/__init__.py	ReMe memory module initialization
src/agentscope/memory/reme/_reme_base_long_term_memory.py	Base class for ReMe memory implementations with context manager support
src/agentscope/memory/reme/_reme_personal_memory.py	Personal memory implementation for user preferences and facts
src/agentscope/memory/reme/_reme_task_memory.py	Task memory implementation for learning from execution trajectories
src/agentscope/memory/reme/_reme_tool_memory.py	Tool memory implementation for recording tool usage patterns
tests/memory_reme_test.py	Comprehensive test suite with dynamic memory type testing
examples/functionality/long_term_memory/reme/*.py	Three example files demonstrating each memory type
examples/functionality/long_term_memory/reme/README.md	Complete documentation with API reference and usage guide
docs/tutorial/zh_CN/src/task_long_term_memory.py	Chinese tutorial integration
docs/tutorial/en/src/task_long_term_memory.py	English tutorial integration
docs/tutorial/_static/css/gallery.css	CSS styling for console output blocks

Comments suppressed due to low confidence (1)

pyproject.toml:21

• The project requires Python 3.10+ (requires-python = '>=3.10'), but reme-ai requires Python 3.12+ (python_full_version >= '3.12'). While this is technically valid since reme-ai is an optional dependency, users may be confused when they cannot install the full extras on Python 3.10-3.11. Consider adding a note in the documentation or README explaining that ReMe memory features require Python 3.12+.

requires-python = ">=3.10"

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Osier-Yi]

LGTM

[DavdGao]

Please see inline comment, others lgtm

[DavdGao]

LGTM