feat: add Mem0 memory integration with config, implementation, docs, tests, and dependency

[kartik-mem0]

Closes #597

What

Adds Mem0 as a new memory store plugin (type: mem0) alongside simple, file, and blackboard.

Changes

File	Change
entity/configs/node/memory.py	Mem0MemoryConfig dataclass with FIELD_SPECS
entity/configs/__init__.py	Export Mem0MemoryConfig
runtime/node/agent/memory/mem0_memory.py	New — Mem0Memory(MemoryBase) using mem0ai SDK
runtime/node/agent/memory/builtin_stores.py	Register mem0 store with lazy factory
pyproject.toml	Add mem0ai>=1.0.9 dependency
tests/test_mem0_memory.py	New — 25 unit tests
yaml_instance/demo_mem0_memory.yaml	New — Example workflow
docs/user_guide/en/modules/memory.md	Mem0 section + comparison table entry

Design

• Uses mem0ai Python SDK with lazy import (no startup cost)
• Search uses filters dict; add uses top-level kwargs (per SDK docs)
• user_id/agent_id scoping with OR filter when both are configured
• Agent output uses role: "assistant" per Mem0's entity-scoped memory model
• load()/save() are no-ops — Mem0 manages persistence server-side
• All errors are caught and logged, never propagated

Test

uv run python -m pytest tests/test_mem0_memory.py -v  # 25 pass
                                                                            

[NA-Wen]

Thanks a lot for this contribution—this is a meaningful enhancement to current agents' memory system and really helps broaden the overall capability.

One small issue we noticed is that the Chinese version of the documentation (docs/user_guide/zh/modules/memory.md) seems to be missing or not updated alongside the English changes. It would be great to keep both language versions in sync to ensure consistency for all users.