Add Cheaper Inference preset

[aiapienthusiast]

This PR adds Cheaper Inference as a preset in the CLI.
It uses the same pattern as #752.

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

Usage:

export CHEAPER_INFERENCE_API_KEY=your-token
humanify cheaperinference obfuscated.js -o readable.js

The default model is gpt-5.4-mini. Use -m to change it.

Details

Files:

• src/cli/cheaperinference.rs: preset defaults and a unit test.
• src/cli/mod.rs, src/main.rs: new cheaperinference subcommand.
• tests/common/mod.rs, tests/e2e_cheaperinference.rs: label-gated e2e test.
• .github/workflows/e2e-cheaperinference.yml: runs with label test-cheaperinference on branches in this repository.
• README.md: usage line, "Cheaper Inference mode" section, e2e label list.

Tests: cargo fmt -- --check, cargo clippy -- -D warnings, cargo test --lib (206 passed), cargo test --test cli_smoke pass.
The e2e test compiles. It is ignored by default, as the other provider e2e tests are.

Live check: gpt-5.4-mini returned a reply through humanify cheaperinference -.
The output had readable names.

The e2e workflow needs a CHEAPER_INFERENCE_API_KEY repository secret to run.