fix(config): generate unique default identifiers

[JiataiWang]

Description

MOSConfig.session_id and GeneralMemCubeConfig.cube_id currently call uuid.uuid4() when their modules are loaded. Pydantic then treats those generated strings as static defaults, so separate config instances created in the same process reuse the same identifier.

This switches both fields to default_factory. Explicit IDs and existing serialized configs keep the same behavior; only omitted IDs now get a fresh UUID per instance.

The PR is now based on dev-v2.0.30; the bug is still present there and the four-file functional diff is unchanged. No dependencies are added. I couldn't find an existing issue or PR for this, and the contribution guide says small bug fixes can go directly to a PR.

Related Issue (Required): N/A

Type of change

• Bug fix (non-breaking change which fixes an issue)

How Has This Been Tested?

• Unit Test
• Test Script Or Test Steps

Verification on the merged dev-v2.0.30 tree:

• uv run --frozen --with 'pytest==8.3.5' pytest tests/configs/ -q — 39 passed
• uv run --frozen --extra mem-reader --with 'pytest==8.3.5' pytest tests/mem_os/ -q — 36 passed (3 existing Pydantic serialization warnings)
• uv run --frozen --with 'poetry>=2,<3' --with 'ruff==0.11.13' make format — Ruff passed; 621 files unchanged
• git diff --check origin/dev-v2.0.30...HEAD — passed
• GitHub Python tests matrix — 16/16 passed across Ubuntu, Windows, macOS 14/15 and Python 3.10–3.13
• Repository autotest — changed tests 4/4 and generated tests 43/43 passed

Checklist

• I have performed a self-review of my own code | 我已自行检查了自己的代码
• I have commented my code in hard-to-understand areas (no hard-to-understand code added) | 我已在难以理解的地方对代码进行了注释
• I have added tests that prove my fix is effective or that my feature works | 我已添加测试以证明我的修复有效或功能正常
• I have created related documentation issue/PR in MemOS-Docs (not applicable; no documentation behavior changed) | 我已在 MemOS-Docs 中创建了相关的文档 issue/PR（如适用）
• I have linked the issue to this PR (not applicable; no existing issue found) | 我已将 issue 链接到此 PR（如适用）
• I have mentioned the person who will review this PR | 我已提及将审查此 PR 的人

@WeiminLee when you have a moment, could you take a look?

Reviewer Checklist

• closes #xxxx (Replace xxxx with the GitHub issue number)
• Made sure Checks passed
• Tests have been provided

[Memtensor-AI]

🤖 Open Code Review

Target: PR #2201
Task: a8b5631b96d328da
Base: dev-v2.0.30
Head: fix/unique-config-identifiers

🔍 OpenCodeReview found 1 issue(s) in this PR.

⚠️ 1 warning(s) occurred during review.

---

1. tests/configs/test_mem_os.py (L5-L6)

Using model_construct() bypasses Pydantic v2's __init__, validators, and default_factory. Because session_id is declared with default_factory=lambda: str(uuid.uuid4()), calling model_construct will not invoke that factory — session_id will simply be absent (or None) on both instances, so first_config.session_id != second_config.session_id evaluates to None != None → False, and the test will fail (or raise AttributeError) rather than proving uniqueness.

Use the normal constructor so that default_factory is executed for each instance:

first_config  = MOSConfig(chat_model=..., mem_reader=...)
second_config = MOSConfig(chat_model=..., mem_reader=...)

See the parallel test in test_mem_cube.py (test_general_mem_cube_config_generates_unique_default_cube_ids) for the correct pattern.

💡 Suggested Change

Before:

    first_config = MOSConfig.model_construct(chat_model=None, mem_reader=None)
    second_config = MOSConfig.model_construct(chat_model=None, mem_reader=None)

After:

    first_config = MOSConfig(chat_model=None, mem_reader=None)
    second_config = MOSConfig(chat_model=None, mem_reader=None)

---

🧹 Filtered 1 low-confidence OCR finding(s) before posting/fix-loop (existing_code_mismatch: 1).

Generated by cloud-assistant via Open Code Review.

[JiataiWang]

Good catch to check. I reran this with the exact test setup: Pydantic v2 model_construct() calls field defaults with call_default_factory=True; both session_id values are strings, parse as UUIDv4, and differ. The same test failed before the fix because both instances received the same import-time UUID. I am using model_construct(chat_model=None, mem_reader=None) only to avoid the invalid empty defaults of the nested configs, not to bypass the session_id factory, so I am leaving the test as-is.

[Memtensor-AI]

⚠️ Automated Test Results: ENV ISSUE

The test environment encountered an issue that requires manual attention.

Details: Both failing tests cannot locate 'examples/data/mem_cube_2/config.json' relative to the current working directory, causing FileNotFoundError before any assertion runs. [advisory, non-gating] AI-generated tests on branch test/auto-gen-a546a3da2f1dfb5b-20260807153108: 66/66 passed — these do NOT affect the PR verdict; review the branch manually.
Branch: fix/unique-config-identifiers

[JiataiWang]

Fixed the CWD dependency in 7e8769c: both tests now resolve the fixture relative to test_mem_cube.py. I reran that file from /private/tmp (3 passed), then the full config suite (39 passed) and mem_os suite (36 passed).

[Memtensor-AI]

✅ Automated Test Results: PASSED

All tests passed (4/4 executed). memos_python_core/changed-repo-python: 4/4. Duration: 5s [advisory, non-gating] AI-generated tests on branch test/auto-gen-2d018f7ca2c29a15-20260808102020: 61/61 passed — these do NOT affect the PR verdict; review the branch manually.

Branch: fix/unique-config-identifiers

[JiataiWang]

Thanks for the approval, @WeiminLee. I noticed the required Python tests workflow is still sitting at action_required and never started any jobs, so GitHub is keeping the PR blocked even though the other checks passed. Could you approve that run when you get a chance?

[Memtensor-AI]

✅ Automated Test Results: PASSED

All tests passed (4/4 executed). memos_python_core/changed-repo-python: 4/4. Duration: 5s [advisory, non-gating] AI-generated tests on branch test/auto-gen-a8b5631b96d328da-20260818094438: 43/43 passed — these do NOT affect the PR verdict; review the branch manually.

Branch: fix/unique-config-identifiers