Feat/tencent tokenhub provider by Contentment003111 · Pull Request #16860 · NousResearch/hermes-agent

Contentment003111 · 2026-04-28T06:45:16Z

Register tencent-tokenhub as a new API-key provider targeting https://tokenhub.tencentmaas.com/v1. Includes full provider wiring (auth, models, doctor, CLI, overlay, aliases), 65 new test cases, and documentation updates.

Files changed: 13 modified, 1 new (+637/-6 lines)

What does this PR do?

Add tencent-tokenhub as a new API-key provider, enabling users to access Tencent's TokenHub MaaS platform (https://tokenhub.tencentmaas.com/v1) through Hermes.

TokenHub is Tencent's model-as-a-service gateway that provides access to models like tencent/hy3-preview. This PR follows the same registration pattern as existing providers (Xiaomi, ArceeAI, etc.) and includes full provider wiring, comprehensive tests, and documentation.

Related Issue

N/A — new provider integration.

Type of Change

🐛 Bug fix (non-breaking change that fixes an issue)
✨ New feature (non-breaking change that adds functionality)
🔒 Security fix
📝 Documentation update
✅ Tests (adding or improving test coverage)
♻️ Refactor (no behavior change)
🎯 New skill (bundled or hub)

Changes Made

Core Implementation

hermes_cli/providers.py — Register tencent-tokenhub in HermesOverlay and _ALIASES
hermes_cli/auth.py — Add ProviderConfig entry in PROVIDER_REGISTRY + aliases (tencent, tokenhub, tencent-cloud, tencentmaas) in resolve_provider()
hermes_cli/models.py — Add default model definitions for tencent-tokenhub
hermes_cli/doctor.py — Include tencent-tokenhub in diagnostic checks
hermes_cli/main.py — Wire up tencent-tokenhub in CLI entry point
agent/model_metadata.py — Add model metadata entries
agent/auxiliary_client.py — Register auxiliary client support

Tests (65 new test cases)

tests/hermes_cli/test_tencent_tokenhub_provider.py — New file (494 lines), 60 tests covering: provider config, auth resolution, alias mapping, model definitions, doctor integration, overlay wiring, model catalog, and env-var isolation
tests/hermes_cli/test_runtime_provider_resolution.py — +5 runtime resolution tests for tencent-tokenhub
tests/hermes_cli/test_xiaomi_provider.py — Add TOKENHUB_API_KEY to env cleanup for hermetic isolation
tests/hermes_cli/test_arcee_provider.py — Add TOKENHUB_API_KEY to env cleanup for hermetic isolation

Documentation

website/docs/integrations/providers.md — Add tencent-tokenhub provider documentation
website/docs/reference/environment-variables.md — Document TOKENHUB_API_KEY
website/static/api/model-catalog.json — Add tencent/hy3-preview to model catalog

How to Test

Run tencent-tokenhub specific tests: pytest tests/hermes_cli/test_tencent_tokenhub_provider.py -v
Run runtime resolution tests: pytest tests/hermes_cli/test_runtime_provider_resolution.py -k "TencentTokenhub" -v
Run all related provider tests to verify no regressions: pytest tests/hermes_cli/test_xiaomi_provider.py tests/hermes_cli/test_arcee_provider.py -v
Verify end-to-end (requires TOKENHUB_API_KEY): export TOKENHUB_API_KEY="your-key-here" && hermes -p tencent-tokenhub -q "Hello"

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 (65 new test cases, 605 total related tests pass)
I've tested on my platform: Linux (CentOS)

Documentation & Housekeeping

I've updated relevant documentation (README, docs/, docstrings)
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

Register tencent-tokenhub as a new API-key provider targeting https://tokenhub.tencentmaas.com/v1. Includes full provider wiring (auth, models, doctor, CLI, overlay, aliases), 65 new test cases, and documentation updates. Files changed: 13 modified, 1 new (+143/-6 lines)

OpenRouter lists the model as tencent/hy3-preview:free (with :free suffix). Update OPENROUTER_MODELS, model-catalog.json, and tests to match.

Registers tencent-tokenhub (https://tokenhub.tencentmaas.com/v1) as a new API-key provider with model tencent/hy3-preview (256K context). - PROVIDER_REGISTRY entry + TOKENHUB_API_KEY / TOKENHUB_BASE_URL env vars - Aliases: tencent, tokenhub, tencent-cloud, tencentmaas - openai_chat transport with is_tokenhub branch for top-level reasoning_effort (Hy3 is a reasoning model) - tencent/hy3-preview:free added to OpenRouter curated list - 60+ tests (provider registry, aliases, runtime resolution, credentials, model catalog, URL mapping, context length) - Docs: integrations/providers.md, environment-variables.md, model-catalog.json Author: simonweng <simonweng@tencent.com> Salvaged from PR #16860 onto current main (resolved conflicts with #16935 Azure Anthropic env-var hint tests and the --provider choices= list removal in chat_parser).

AUTHOR_MAP entry for the tencent-tokenhub provider PR #16860 contributor.

teknium1 · 2026-04-28T10:46:15Z

Merged via #16960 — your commit was cherry-picked onto current main with your authorship preserved (see commit a6a6cf0). Thanks for the thorough provider integration and the 60+ tests!

Salvage notes: resolved a conflict with #16935 in the runtime-resolver tests (kept both test classes), and dropped a stray blank line in agent/model_metadata.py. The --provider choices= list in chat_parser was recently removed on main (user-defined providers from providers: in config.yaml are now valid), so that part of your diff was dropped in favor of main's newer approach.

Registers tencent-tokenhub (https://tokenhub.tencentmaas.com/v1) as a new API-key provider with model tencent/hy3-preview (256K context). - PROVIDER_REGISTRY entry + TOKENHUB_API_KEY / TOKENHUB_BASE_URL env vars - Aliases: tencent, tokenhub, tencent-cloud, tencentmaas - openai_chat transport with is_tokenhub branch for top-level reasoning_effort (Hy3 is a reasoning model) - tencent/hy3-preview:free added to OpenRouter curated list - 60+ tests (provider registry, aliases, runtime resolution, credentials, model catalog, URL mapping, context length) - Docs: integrations/providers.md, environment-variables.md, model-catalog.json Author: simonweng <simonweng@tencent.com> Salvaged from PR NousResearch#16860 onto current main (resolved conflicts with NousResearch#16935 Azure Anthropic env-var hint tests and the --provider choices= list removal in chat_parser).