Packages

MCP server that enables AI agents to interact with macOS through AppleScript, providing tools for controlling Calendar, Clipboard, Finder, Notifications, System settings, iTerm, Shortcuts, Mail, Messages, Notes, and Pages.

Unified caching library for Python — provides a consistent interface across multiple cache backends (Redis, Memcached, filesystem, SimpleCache in-memory). cachelib features: SimpleCache (dict-based in-memory), FileSystemCache (disk-based with pickle), RedisCache (via redis-py), MemcachedCache (via pylibmc/python-memcached), SASLMemcachedCache, UWSGICache, NullCache (for testing), uniform API (set/get/delete/clear/has), timeout support, key prefixing, and threshold-based eviction for SimpleCache. Flask's cache backend (used by Flask-Caching extension).

Async connection pool library for Rust with built-in backends for PostgreSQL (via tokio-postgres), Redis, Memcached, and generic pool support. Provides async-aware pooling with configurable pool size, connection lifetime, health checks, and reconnection. More ergonomic than bb8 for PostgreSQL-specific use cases. Used in Rust web services alongside Actix-web, Axum, and other async frameworks.

Java/Kotlin implementation of gRPC — Google's high-performance RPC framework using Protocol Buffers over HTTP/2. Supports unary, server-streaming, client-streaming, and bidirectional streaming RPCs. Generates Java stubs from .proto files with the protoc plugin. The standard for building high-performance agent microservice communication in JVM ecosystems, with built-in load balancing, retries, and deadline propagation.

Modern incident management platform with Slack-native workflows, automated escalation, runbooks, and postmortem automation for engineering teams.

Thread-safe persistent queue library for Python — provides disk-backed queues that survive process restarts using SQLite or filesystem storage. persistqueue features: Queue (SQLite-backed FIFO), FIFOSQLiteQueue/UniqueQ/FIFOUniqueQ variants, SQLiteAckQueue (explicit acknowledgment for at-least-once delivery), FIFOSQLiteAckQueue, put()/get()/task_done()/ack()/nack() API, auto-commit mode, serialization support (pickle, msgpack, JSON), thread-safe operations, SQLite WAL mode for concurrent access, and empty/qsize/full checks. Drop-in for Python's queue.Queue but with disk persistence.

Out-of-core lazy dataframe library for Python — processes billion-row datasets without loading into RAM using memory-mapped files. vaex features: open() for memory-mapped HDF5/Arrow/CSV, lazy evaluation (computations deferred until needed), virtual columns (computed on-the-fly), df.mean/std/sum operations at billion rows/second, df.plot1d/plot2d for fast statistical plots, filtering with boolean expressions, df.apply() for UDFs, df.export_hdf5() for efficient storage, string operations, ML feature engineering, and JIT compilation via Numba/Pytables. Enables desktop analysis of datasets too large for pandas.

Lightweight agent framework where every agent is a composable unit with explicit Pydantic input/output schemas, enabling chainable, independently testable agent pipelines with minimal dependencies.

Delivers the fastest publicly available LLM inference (2,000+ tokens/second on Llama 3 models) using NVIDIA WSE-3 wafer-scale chips, with an OpenAI-compatible REST API and a currently limited but growing model catalog.

JetBrains extension of Jetpack Compose for Kotlin Multiplatform. Write shared UI code in Compose that runs on Android, iOS, Desktop (JVM), and Web (Wasm). Share not just business logic but also UI components across platforms. Uses the same @Composable function model as Jetpack Compose — Android Compose knowledge transfers directly. iOS support reached stable status in 2024.

Esri ArcGIS Platform REST API for enterprise geographic information systems (GIS) and geospatial intelligence platform. Enables AI agents to manage geospatial data layer publishing and access control, handle spatial query and analysis automation using geocoding, routing, and proximity services, access feature service CRUD operations for geographic data management, retrieve map and scene service data for visualization and analysis, manage web map and dashboard configuration automation, handle geocoding and reverse geocoding for address and location data, access geoprocessing service orchestration for spatial workflows, retrieve network analysis data for routing, service areas, and logistics optimization, manage portal and organization content and permission management, and integrate geospatial data with enterprise asset management, field operations, and analytics platforms.

Ultra-fast LLM inference API powered by custom LPU (Language Processing Unit) hardware delivering 300-500+ tokens/second for Llama, Mixtral, and Gemma models.

Open-source LLM observability platform — traces, evaluates, and monitors LLM applications with a REST API, Python/JS SDKs, and native OpenAI/LangChain/CrewAI integrations.

Markdown to WeChat MCP server enabling AI agents to publish content to WeChat Official Accounts (微信公众号) — converting Markdown content to WeChat's rich text format, uploading images, drafting articles in WeChat backend, and integrating WeChat Official Account content publishing into agent-driven content creation workflows.

JIT compiler for Python that translates Python/NumPy code to native machine code using LLVM. Add @jit or @njit decorator to compute-intensive Python functions to achieve C/Fortran-like performance. Supports NumPy array operations, loops, and CUDA GPU acceleration via @cuda.jit. Used when vectorized NumPy operations don't fully remove Python overhead from hot loops.

Python ORM library that combines SQLAlchemy and Pydantic in a single model definition. Define a SQLModel class once and use it as both a database table model (SQLAlchemy) and an API schema (Pydantic). Designed by FastAPI's creator (Sebastián Ramírez) to eliminate the need for separate ORM and API schema models in FastAPI applications.

SVG to PNG/PDF/PS/EPS converter for Python — uses Cairo graphics library to render SVG files to raster and vector formats. cairosvg features: svg2png(), svg2pdf(), svg2ps(), svg2eps() conversion functions, file or bytes input/output, DPI control, scale factor, background color, custom width/height, URL stylesheet support, and full SVG 1.1 specification support including CSS, filters, gradients, and transforms. Requires Cairo C library installed on system.

Official Go implementation of gRPC — Google's high-performance RPC framework using Protocol Buffers over HTTP/2. Generates Go server and client code from .proto definitions using protoc-gen-go-grpc. Supports all four gRPC RPC types: unary, server-streaming, client-streaming, and bidirectional streaming. Includes interceptors for auth, logging, and metrics. The standard for Go microservice communication.

Simple, blocking HTTP client for Rust. Zero async complexity — synchronous API with no tokio/async-std dependency. Small binary footprint. Supports TLS (via rustls or native-tls), JSON (via serde_json feature), redirects, and timeouts. Perfect for CLI tools, scripts, and simple applications that don't need async HTTP.

Kubernetes progressive delivery controller providing advanced deployment strategies (canary, blue-green) with automated analysis and rollback. Argo Rollouts extends Kubernetes Deployment with traffic shaping (via Istio, nginx, ALB, Contour), metric analysis (Prometheus, Datadog, New Relic), and automated promotion/rollback based on success metrics. Enables zero-downtime deployments with automatic rollback when error rates spike. Part of the Argo project (alongside ArgoCD, Argo Workflows, Argo Events).