Packages

MCP server enabling AI agents to search for flights using travel data APIs (Amadeus, Google Flights via SerpAPI, or similar). Enables agents to find available flights, compare prices, check availability, and retrieve flight schedule information to support travel planning workflows.

Matrix protocol MCP server enabling AI agents to interact with Matrix-based decentralized messaging — sending and reading messages in Matrix rooms, managing room membership, and integrating Matrix's open federated chat protocol into agent-driven communication and community management workflows.

Most popular Java mocking framework for unit testing. Creates mock objects, stubs method calls (when().thenReturn()), and verifies interactions (verify()). Works with JUnit 5 via @ExtendWith(MockitoExtension.class) or Spring's @MockBean. Mockito's fluent API makes test setup readable and concise — the standard for Java unit testing with mocks.

Mocking library for .NET — creates test doubles (mocks) of interfaces and abstract classes using lambda-based LINQ setup API. Moq features: var mock = new Mock<IAgentRepository>(); mock.Setup(r => r.FindAsync(agentId)).ReturnsAsync(agent) stubs method return; mock.Verify(r => r.SaveAsync(It.IsAny<Agent>()), Times.Once()) verifies call occurred; Mock.Of<T>() for inline mock creation; MockBehavior.Strict throws on unexpected calls; Callback for capturing arguments; SetupSequence for multiple sequential returns. Moq uses Castle.DynamicProxy to generate mock implementations at runtime. Most widely used .NET mocking library. Works with xUnit, NUnit, and MSTest for agent service unit tests.

Mock library for Elixir that enforces behaviour-based contracts — creates test mocks that validate against Elixir @behaviour callbacks. Mox features: Mox.defmock/2 creates mock modules, expect/4 sets expectations with call count, stub/3 allows any number of calls, verify_on_exit!/1 asserts all expectations satisfied, allow/3 for async test access, global mode for non-concurrent tests, and strict adherence to @behaviour interface. Created by José Valim (Elixir creator) with explicit-contract philosophy — mocks must match a defined behaviour.

Open-source hyperparameter optimization (HPO) framework using a 'define-by-run' API. Unlike grid search or random search, Optuna uses efficient algorithms (TPE, CMA-ES) to intelligently explore the hyperparameter space. Supports distributed optimization across multiple machines, pruning (early stopping of unpromising trials), and integration with any ML framework (PyTorch, TensorFlow, XGBoost, etc.). Also used for general black-box optimization beyond ML hyperparameters.

The fastest and most efficient pagination Ruby gem — 40x faster and 100x more memory-efficient than kaminari or will_paginate. Pagy doesn't pollute models with pagination logic; works at the controller level. Key API: pagy(collection) returns [pagy, records] tuple; pagy_nav(pagy) renders pagination links in views. Supports: multiple pagination UI styles (bootstrap, bulma, tailwind, responsive), JSON API pagination metadata, cursor-based pagination (pagy-keyset for cursor), elasticsearch, searchkick, meilisearch integration. Pagy extras provide optional features without loading unused code. Rails standard includes pagy_metadata for JSON API pagination responses.

Python password hashing library providing a unified interface to 30+ password hashing algorithms (bcrypt, argon2, scrypt, pbkdf2, sha256_crypt, etc.). Handles hashing, verification, and hash migration. Used in FastAPI and Flask user authentication systems for secure password storage. The standard Python password hashing library despite being in maintenance mode.

Fast DataFrame library for Python written in Rust with Apache Arrow columnar memory format. 5-100x faster than pandas for many operations. Lazy evaluation API for query optimization. Strong type system with no null index. Multi-threaded execution. Handles out-of-memory datasets via streaming. Growing alternative to pandas for performance-sensitive data pipelines.

High-performance DataFrame library for Python and Rust, built on Apache Arrow and written in Rust. 10-100x faster than pandas for most operations due to lazy evaluation, query optimization, and multi-threaded execution. Supports both eager and lazy API, SQL queries, streaming for out-of-core data, and Parquet/CSV/JSON I/O. The modern pandas replacement for data-intensive agent pipelines.

A declarative JavaScript UI component library for building interactive user interfaces using a component model, hooks, and (in React 19) Server Components.

Solana MCP server enabling AI agents to interact with the Solana blockchain — querying account balances, fetching transaction history, reading on-chain program data, monitoring token accounts, accessing DeFi protocol state, and integrating Solana blockchain data into agent-driven crypto analytics, portfolio management, and Web3 development workflows.

Modern CSS linter that catches errors and enforces style conventions in CSS, SCSS, Sass, Less, and CSS-in-JS. 170+ built-in rules for property ordering, color format consistency, selector specificity, and syntax errors. Integrates with editors (VS Code, JetBrains) and CI pipelines. Stylelint v16 dropped Node 14/16 support and moved to flat config format.

Utility-first CSS framework that provides low-level CSS utility classes (text-lg, flex, p-4, bg-blue-500) to build custom designs without writing CSS. Tailwind's JIT compiler generates only the CSS classes used in your code, keeping production CSS tiny. v4 moved to CSS-native configuration. Widely used with React, Vue, Next.js, and any modern frontend stack. The dominant CSS framework as of 2024.

Python framework for building sophisticated Terminal User Interfaces (TUIs). Textual brings CSS-like styling, reactive components, layout engines, and an event system to terminal applications. Built on top of Rich, it enables building full-screen terminal apps that feel like modern web UIs but run entirely in the terminal. Used for agent monitoring dashboards, CLI tools, and developer tools.

Official client-side router for Vue.js — maps URL paths to Vue components with support for nested routes, dynamic segments, navigation guards, and lazy loading. Vue Router 4 features: createRouter() + createWebHistory()/createWebHashHistory(), route params (/agents/:id via useRoute().params.id), nested routes with children[] array, navigation guards (beforeEach, beforeRouteEnter, beforeRouteUpdate), lazy loading (component: () => import('./AgentView.vue')), route meta for auth checks, programmatic navigation (useRouter().push('/agents')), named routes, route transitions, scroll behavior, and 404 catch-all routes. Deeply integrated with Vue 3 Composition API and Vue DevTools.

YouTube's data API for searching videos, retrieving channel/video metadata, managing playlists, and accessing captions and comments programmatically.

Blazing-fast structured logging library for Go, created by Uber. Provides two loggers — zap.NewProduction() for JSON structured logs with field typing, and zap.NewExample() / zap.NewDevelopment() for development. Fields are typed (zap.String, zap.Int, zap.Duration) avoiding reflection overhead. 10-100x faster than standard Go log package for high-throughput services.

Lightweight, fast Node.js library for adding ANSI color and style codes to terminal strings. Used by Gulp.js, Mocha, and many popular CLI tools. Zero dependencies, CommonJS compatible, and faster than chalk in many benchmarks. Supports colors, bright colors, background colors, and text styles. Respects NO_COLOR and FORCE_COLOR environment variables.

Express.js middleware for configuring Cross-Origin Resource Sharing (CORS) headers. Handles the browser's same-origin policy by setting Access-Control-Allow-Origin, Access-Control-Allow-Methods, Access-Control-Allow-Headers, and related CORS headers on HTTP responses. Supports whitelist-based origin validation, preflight OPTIONS request handling, credential cookies, and dynamic per-request origin configuration. Part of the expressjs GitHub organization.