```js const mocks = { 'get /api/message': { message: 'hello' }, 'post /api/model/:id': (req, res) => { res.send({ id: req.params.id }); } };

Microsoft Azure Functions NodeJS Framework

Default model selection shared by Agent entry points

Route Model Binding. When you inject a model ID into a route, you will often look to retrieve the model that corresponds to that ID. This package provides a convenient solution for Adonis.

Easily display interactive 3D models on the web and in AR!

Generic vision proxy extension for pi: lets non-multimodal models (e.g. DeepSeek) delegate image analysis to ANY OpenAI-compatible multimodal model. Zero vendor hardcoding - a model is just baseUrl + apiKey + model id. Auto-hides itself when the main mode

Core types and registry utilities for TokenLens (model metadata).

Model-facing ask_user_question tool over the ctx.userInteraction seam

TUF metadata models

Model-facing owner-scoped persistent Bash tool backed by the Harness PTY service

The **[Fireworks provider](https://ai-sdk.dev/providers/ai-sdk-providers/fireworks)** for the [AI SDK](https://ai-sdk.dev/docs) contains language model and image model support for the [Fireworks](https://fireworks.ai) platform.

TensorFlow layers API in JavaScript

Types and validators that help describe the model of a Backstage Catalog

JSON becomes real things. Define your catalog, register your components, let AI generate.

This package contains helpers to process webpack stats

Model-facing workflow tool: run a JavaScript orchestration script over ctx.workflows

An implementation of the CSSStyleDeclaration class from the CSS Object Model specification

Model-facing owner-scoped persistent PowerShell tool backed by the Harness PTY service

In-process spawn subagent backend: runs a fresh child agent on ctx.agents

JSON file KV storage backend for the DeepSeek Harness storage hub