feat: add Cheaper Inference as an LLM provider

[aiapienthusiast]

This PR adds Cheaper Inference as an LLM provider preset in llm_client.py.
It uses the same pattern as #39 (OrcaRouter).

Cheaper Inference is an OpenAI-compatible LLM gateway.
One API key gives access to models from several labs.

Usage:

[api_keys]
cheaperinference_api_key = "ci_live_..."

[llm]
provider = "cheaperinference"
model = "gpt-5.4-mini"

You can also set CHEAPER_INFERENCE_API_KEY instead of the TOML key.

Details

Files:

• src/standard/llm_client.py: new cheaperinference entry in PROVIDER_DEFAULTS (base URL https://api.cheaperinference.com/v1, default model gpt-5.4-mini) and the CHEAPER_INFERENCE_API_KEY env mapping.
• tests/test_llm_client.py: provider defaults test and env key test.
• config/config.example.toml, README.md, README-zh.md, docs/configuration.md, docs/api-reference.md, docs/pipeline.md: config example and provider lists.

Tests: python -m pytest -q passes on Python 3.10 and 3.12 (42 passed).

Live check: gpt-5.4-mini returned a reply through resolve_llm_config() and chat_completions() with the default temperature 1.3.

Docs: https://cheaperinference.com/docs
Models: https://cheaperinference.com/markets