feat: add MiniMax as configurable LLM provider for summarization and code interpreter

[octo-patch]

Summary

• Makes the LLM provider configurable via LLM_PROVIDER env var (openai or minimax)
• utils/summarise.py: when LLM_PROVIDER=minimax, uses AsyncOpenAI pointed at https://api.minimax.io/v1 with model MiniMax-M2.7 (204K context window)
• service/code_interpreter.py: same provider routing; temperature is clamped to 0.01 for MiniMax because the API requires temperature > 0
• .env.example: documents the two new env vars (LLM_PROVIDER, MINIMAX_API_KEY)
• README.md: adds a "LLM providers" table listing OpenAI (default) and MiniMax
• tests/: 17 unit tests (fully mocked) + 2 integration tests against the real MiniMax API

Test plan

• Unit tests: ============================= test session starts ==============================
  platform linux -- Python 3.10.12, pytest-8.2.0, pluggy-1.6.0
  codspeed: 4.3.0 (disabled, mode: walltime, callgraph: not supported, timer_resolution: 1.0ns)
  benchmark: 5.2.3 (defaults: timer=time.perf_counter disable_gc=False min_rounds=5 min_time=0.000005 max_time=1.0 calibration_precision=10 warmup=False warmup_iterations=100000)
  rootdir: /home/ximi/github_pr/automation/repos/super-rag
  configfile: pyproject.toml
  plugins: logfire-4.29.0, flaky-3.8.1, syrupy-5.1.0, recording-0.13.4, anyio-4.13.0, order-1.3.0, examples-0.0.18, mock-3.15.1, timeout-2.4.0, asyncio-0.23.6, respx-0.22.0, cov-4.1.0, hydra-core-1.3.2, benchmark-5.2.3, typeguard-4.5.1, codspeed-4.3.0, langsmith-0.7.22, dotenv-0.5.2, socket-0.7.0, requests-mock-1.12.1, Faker-37.12.0, httpx-0.36.0, xdist-3.8.0
  asyncio: mode=strict
  collected 19 items / 2 deselected / 17 selected

tests/test_minimax_provider.py ................. [100%]

=============================== warnings summary ===============================
tests/test_minimax_provider.py:270
/home/ximi/github_pr/automation/repos/super-rag/tests/test_minimax_provider.py:270: PytestUnknownMarkWarning: Unknown pytest.mark.integration - is this a typo? You can register custom marks to avoid this warning - for details, see https://docs.pytest.org/en/stable/how-to/mark.html
@pytest.mark.integration

tests/test_minimax_provider.py::TestGenerateContent::test_contains_document_content
/home/ximi/github_pr/automation/repos/super-rag/models/document.py:72: PydanticDeprecatedSince20: Pydantic V1 style @validator validators are deprecated. You should migrate to Pydantic V2 style @field_validator validators, see the migration guide for more details. Deprecated in Pydantic V2.0 to be removed in V3.0. See Pydantic V2 Migration Guide at https://errors.pydantic.dev/2.12/migration/
@validator("id")

tests/test_minimax_provider.py::TestGenerateContent::test_contains_document_content
/home/ximi/github_pr/automation/repos/super-rag/models/document.py:80: PydanticDeprecatedSince20: Pydantic V1 style @validator validators are deprecated. You should migrate to Pydantic V2 style @field_validator validators, see the migration guide for more details. Deprecated in Pydantic V2.0 to be removed in V3.0. See Pydantic V2 Migration Guide at https://errors.pydantic.dev/2.12/migration/
@validator("dense_embedding")

tests/test_minimax_provider.py::TestCompletion::test_completion_openai
/home/ximi/.local/lib/python3.10/site-packages/pytest_asyncio/plugin.py:759: DeprecationWarning: The event_loop fixture provided by pytest-asyncio has been redefined in
/home/ximi/github_pr/automation/repos/super-rag/tests/conftest.py:7
Replacing the event_loop fixture with a custom implementation is deprecated
and will lead to errors in the future.
If you want to request an asyncio event loop with a scope other than function
scope, use the "scope" argument to the asyncio mark when marking the tests.
If you want to return different types of event loops, use the event_loop_policy
fixture.

warnings.warn(

-- Docs: https://docs.pytest.org/en/stable/how-to/capture-warnings.html
================ 17 passed, 2 deselected, 4 warnings in 50.00s ================= — 17 passed

• Integration tests: — 2 passed
• Existing behaviour unchanged when LLM_PROVIDER is unset (defaults to OpenAI)
• Set LLM_PROVIDER=minimax + MINIMAX_API_KEY in .env, run the ingest → query pipeline and verify summaries are generated via MiniMax