feat: add MiniMax as first-class LLM provider (M2.7/M2.5, 204K context)

[octo-patch]

Summary

This PR adds MiniMax as a first-class LLM provider for Aser via a new aser/providers.py module.

MiniMax offers OpenAI-compatible API access to its latest model families, all with 204K-token context windows — the largest available in this class.

Changes

File	Description
aser/providers.py	New module: MINIMAX config preset, MINIMAX_MODELS dict (MiniMax-M2.7, MiniMax-M2.7-highspeed, MiniMax-M2.5, MiniMax-M2.5-highspeed), OPENAI preset, custom_provider() helper
aser/__init__.py	Export the providers submodule
examples/agent_minimax.py	End-to-end example: basic agent, high-speed variant, inline config
tests/test_providers.py	26 unit tests + 3 live integration tests (all passing)
.env.example	MINIMAX_API_KEY entry with model notes
README.md / README_CN.md	"Supported LLM Providers" table + MiniMax quick-start code snippet

Quick start

export MINIMAX_API_KEY=your_api_key_here

from aser.agent import Agent
from aser.providers import MINIMAX, MINIMAX_MODELS

agent = Agent(
    name="minimax_agent",
    model=MINIMAX_MODELS["standard"],   # "MiniMax-M2.7"
    model_config=MINIMAX,
)
response = agent.chat("What is MiniMax AI?")
print(response)

Supported models

Key	Model	Context
standard	MiniMax-M2.7	204K
fast	MiniMax-M2.7-highspeed	204K
standard_v25	MiniMax-M2.5	204K
fast_v25	MiniMax-M2.5-highspeed	204K

Tests

pytest tests/test_providers.py -v
29 passed in 11.21s