@playrunner/huggingface

Hugging Face Inference Providers execution for Playrunner workflows.

View source on GitHub

Install

npm install @playrunner/huggingface

Add the package as a direct production dependency of the Playrunner frontend and orchestrator. Each build discovers the contribution declared for its surface.

Package surfaces

• @playrunner/huggingface exports the Hugging Face integration, configuration panel, settings UI, and icon URL.
• @playrunner/huggingface/orchestrator exports the Inference Providers executor.
• @playrunner/huggingface/e2e exports the package-owned E2E contribution.
• @playrunner/huggingface/assets/huggingface.svg exports the package-owned icon asset.

import huggingFaceIntegration, {
  HuggingFaceConfigPanel,
  HuggingFaceSettingsModal,
  huggingFaceIconUrl,
} from '@playrunner/huggingface';
import huggingFaceOrchestratorContribution from '@playrunner/huggingface/orchestrator';

Testing

Run the package checks and unit tests from the repository root:

npm run format:check --prefix packages/huggingface
npm run lint --prefix packages/huggingface
npm run typecheck --prefix packages/huggingface
npm test --prefix packages/huggingface
npm run test:e2e:mock -- --grep @huggingface

The E2E scenario persists and removes a Hugging Face access token. It runs in deterministic mock mode through the shared Playrunner browser harness, which uses the real frontend, API, authentication, and dedicated E2E database. No Hugging Face credentials or live inference requests are required.

Documentation

See the Hugging Face integration documentation for token setup, task configuration, outputs, exports, and Playrunner build integration.

License

Licensed under the Playrunner Sustainable Use License.