fix(models): add live model fetch for tencent-tokenhub provider

[liuhao1024]

What does this PR do?

Adds a live model fetch block for the tencent-tokenhub provider in provider_model_ids(). Previously, the provider only returned the static fallback list containing hy3-preview, even though the TokenHub API (https://tokenhub.tencentmaas.com/v1/models) returns 58 available models.

The fix follows the same pattern used by other API-key providers (stepfun, gmi): resolve credentials via resolve_api_key_provider_credentials(), then call the standard fetch_api_models() helper to query the /v1/models endpoint.

Related Issue

Fixes #59694

Type of Change

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

Changes Made

• hermes_cli/models.py: Added a tencent-tokenhub live-fetch block in provider_model_ids() (13 lines) that queries the TokenHub /v1/models endpoint when credentials are available.

How to Test

1. Configure TOKENHUB_API_KEY in .env (or via hermes auth add tencent-tokenhub)
2. Run hermes model and select the tencent-tokenhub provider
3. Observed result: All 58 models from the TokenHub API appear in the picker (hy3, kimi-k2.7-code-highspeed, glm-5.2, minimax-m3, deepseek-v4-pro-202606, qwen3.5-plus, etc.)

Without the fix, only hy3-preview appears in the picker.

Checklist

Code

• I've read the Contributing Guide
• My commit messages follow Conventional Commits (fix(scope):, feat(scope):, etc.)
• I searched for existing PRs to make sure this isn't a duplicate
• My PR contains only changes related to this fix/feature (no unrelated commits)
• I've run pytest tests/ -q and all tests pass
• I've added tests for my changes (required for bug fixes, strongly encouraged for features)
• I've tested on my platform: macOS 15.2

Documentation & Housekeeping

• I've updated relevant documentation (README, docs/, docstrings) — or N/A
• I've updated cli-config.yaml.example if I added/changed config keys — or N/A
• I've updated CONTRIBUTING.md or AGENTS.md if I changed architecture or workflows — or N/A
• I've considered cross-platform impact (Windows, macOS) per the compatibility guide — or N/A
• I've updated tool descriptions/schemas if I changed tool behavior — or N/A

[alt-glitch]

This was generated by AI during triage.

Fix PR for #59694 (tencent-tokenhub shows only 1 of 58 models). Related to #57500 (adds a new tencent-tokenplan provider + the hy3 model) — distinct scope; that PR does not add this live-fetch block. Not a duplicate.

[teknium1]

Thanks for the focused fix. The premise remains valid on current main: provider_model_ids() only reaches generic live discovery when get_provider_profile() returns a profile (hermes_cli/models.py:2551-2563), while tencent-tokenhub has no bundled model-provider profile and therefore falls through to its one-item static catalog (hermes_cli/models.py:394-396, 2602-2605).

Problems

• The PR diff changes only hermes_cli/models.py; it does not add coverage for the new live-fetch branch or fallback. Existing TokenHub catalog tests begin at tests/hermes_cli/test_tencent_tokenhub_provider.py:147, while the analogous live/fallback pattern is covered for GMI at tests/hermes_cli/test_gmi_provider.py:87-128.

Suggested changes

• Add mocked TokenHub live-success and probe-failure fallback tests alongside its existing provider tests, following the GMI pattern.

Automated hermes-sweeper review.