Packages

Open-source Node.js headless commerce engine with modular architecture for building custom e-commerce backends with full API control.

Python video editing library built on FFmpeg. MoviePy provides a Pythonic API for cutting, concatenating, adding effects, compositing video clips, extracting audio, and rendering text/image overlays on video. Used for automated video generation pipelines, thumbnail creation, video summarization tools, and anywhere programmatic video manipulation is needed.

Open-source authentication library for Next.js and other frameworks (SvelteKit, SolidStart, Express). Provides OAuth/OIDC integration with 50+ providers (Google, GitHub, Discord, etc.), magic links, credentials auth, and JWT/database sessions. Auth.js v5 is the framework-agnostic rewrite of NextAuth.js v4 — works in Edge Runtime, server components, and all modern JS frameworks. The most popular Next.js authentication solution.

Numerical computing library for Elixir that provides multi-dimensional tensor operations with NumPy-like APIs. Nx supports multiple backends: EXLA (XLA/GPU accelerated), TorchX (LibTorch), and binary/BinaryBackend (pure Elixir). Core of the Elixir ML ecosystem — Axon (neural networks), Bumblebee (HuggingFace models), and Scholar (classical ML) are all built on Nx. Enables ML inference and data processing in Elixir without leaving the BEAM runtime.

Open Exchange Rates provides a JSON REST API delivering real-time and historical currency exchange rates for 200+ currencies, updated hourly, using a USD base on the free tier.

Real-time, bidirectional, event-based communication library for Node.js and browsers. Socket.IO provides WebSocket connections with automatic fallback to polling, rooms and namespaces for message grouping, reconnection handling, and acknowledgements. The de facto standard for real-time web applications (chat, live updates, collaborative tools, gaming).

Open-source, Google Zanzibar-inspired fine-grained authorization system. SpiceDB stores permission relationships in a schema and evaluates permission checks by traversing relationship graphs. Implements consistent, scalable permissions checks at the tuple level — enabling permissions like 'user:alice can view document:doc123 because she's in group:engineering which has viewer access to project:backend'. Built by AuthZed.

Spring's Enterprise Integration Patterns (EIP) framework — provides message channels, transformers, routers, filters, service activators, and adapters for building integration flows. Spring Integration implements patterns from Gregor Hohpe's 'Enterprise Integration Patterns' book. Core concepts: Message (header + payload), MessageChannel (pipe between components), MessageHandler (process/transform messages), Adapters (inbound/outbound for Kafka, RabbitMQ, AMQP, JMS, FTP/SFTP, HTTP, TCP, JDBC, WebSocket, Redis, AWS S3/SQS, email). Java DSL: IntegrationFlow.from(source).transform().filter().handle(handler).get(). Used for data integration pipelines, event-driven workflows, and system-to-system messaging.

The Terraform Cloud and Enterprise REST API lets agents programmatically manage workspaces, trigger and monitor plan/apply runs, read and write state, configure variables, and manage organizations and teams.

V7 Labs REST API for AI training data annotation and dataset management platform. Enables AI agents to manage dataset creation and image, video, and document annotation workflows, handle annotation job assignment and quality review automation, access model-assisted labeling and auto-annotation result retrieval, retrieve annotation export in COCO, YOLO, Pascal VOC, and custom ML formats, manage annotation team and reviewer assignment workflows, handle consensus and disagreement detection for annotation quality, access Darwin dataset versioning and lineage tracking, retrieve annotation statistics and labeling progress analytics, manage webhook-based annotation completion notifications, and integrate training data with PyTorch, TensorFlow, Hugging Face, and MLOps pipeline platforms.

MCP server for Weibo — China's dominant microblogging and social media platform. Enables AI agents to read Weibo posts, search content, retrieve user profiles, and monitor trending topics on Weibo — supporting social media monitoring and content analysis workflows for Chinese market teams.

Setup guides and helper scripts for connecting Burp Suite's MCP Server plugin to multiple AI backends (Codex CLI, Ollama, Gemini CLI, LM Studio). Provides 8 specialized security analysis prompt templates for passive vulnerability analysis of real captured traffic — not automated scanning.

High-performance RPC framework for Python — generates client/server stubs from Protocol Buffer definitions for typed, efficient inter-service communication. gRPC Python features: grpc.server() and stub classes from protoc-generated code, unary/server-streaming/client-streaming/bidirectional streaming RPCs, grpc.secure_channel() with TLS, metadata for headers, interceptors for middleware, health checking (grpc_health_checking), reflection, grpc.aio for async support, channel credentials and call credentials, deadlines/timeouts, and grpc.StatusCode error codes. Standard choice for agent-to-agent and microservice communication.

PyTorch Image Models — largest collection of pretrained computer vision models for PyTorch. timm features: 1000+ pretrained models (ViT, EfficientNet, ConvNeXt, Swin Transformer, DeiT, EVA, MetaFormer), timm.create_model() factory with pretrained=True, feature extraction (features_only=True), custom classifier heads, timm.data.create_transform for optimal preprocessing per model, model listings (timm.list_models()), HuggingFace Hub integration (timm.create_model('hf-hub:timm/model')), and benchmark data for model selection. Maintained by Ross Wightman at Hugging Face. Premier library for vision transfer learning in PyTorch.

Fast, low-level WebSocket server and client library for Node.js implementing the standard WebSocket protocol with ping/pong heartbeat and binary data support.

Real-time messaging infrastructure providing WebSocket, MQTT, and SSE transports with guaranteed message ordering, persistence, and 99.999% uptime SLA.

Provides automated visual regression testing and UI review for Storybook component libraries by capturing pixel-level snapshots, detecting visual diffs, and managing human approval workflows through a SaaS platform.

Provides S3-compatible object storage with zero egress fees, integrating natively with Cloudflare Workers and supporting standard AWS S3 SDKs pointed at a custom endpoint.

LLM evaluation framework with 14+ built-in metrics (hallucination, answer relevancy, bias, toxicity, etc.) that integrates with pytest for CI/CD pipelines and optionally syncs results to the Confident AI cloud dashboard.

JetBrains' SQL ORM library for Kotlin — provides two APIs: DSL (type-safe Kotlin SQL DSL) and DAO (ActiveRecord-style entity objects). The DSL API maps Kotlin operations to SQL: SchemaUtils.create(Users), Users.select { Users.name eq "agent" }.firstOrNull(). The DAO API provides Entity classes (similar to Hibernate) with automatic CRUD. Supports PostgreSQL, MySQL, SQLite, Oracle, H2, SQL Server, and MariaDB. Designed for Kotlin-first usage (coroutine support, Kotlin data types). The go-to database library for Ktor backend services.