Packages

Fully managed data pipeline service that automates data movement from 500+ sources to data warehouses with zero-maintenance connectors, offering a REST API for connector and sync management automation.

Autodesk Fusion 360 MCP server enabling AI agents to interact with Fusion 360 — Autodesk's integrated CAD/CAM/CAE cloud platform for product design and manufacturing. Enables creating and modifying parametric 3D designs, running simulations, generating toolpaths for CNC machining, and integrating AI assistance into engineering design workflows.

MCP server for searching the public MCP Registry — enabling AI agents to discover MCP servers, find servers for specific tasks, and copy SSE/HTTP URLs for integration. A meta-tool that makes the MCP ecosystem searchable from within AI agent workflows.

MCP server providing access to Process Hacker — the powerful open-source Windows process and memory monitoring tool (similar to Sysinternals Process Monitor). Enables AI agents to query running Windows processes, inspect process memory, analyze network connections, and monitor system handles through Process Hacker's API.

Reddit MCP server enabling AI agents to interact with Reddit — reading posts and comments from subreddits, searching Reddit content, retrieving user profiles and post history, monitoring discussions, and integrating Reddit's community content platform into agent-driven market research, sentiment analysis, and community monitoring workflows.

Official Rohlik MCP server enabling AI agents to interact with Rohlik — the Czech/Central European online grocery delivery service — searching products, managing shopping carts, placing orders, tracking deliveries, and integrating grocery shopping automation into agent-driven household management and meal planning workflows.

Snowflake's SQL API and Python connector for executing SQL queries against Snowflake's cloud data platform. Supports synchronous and asynchronous query execution, DDL/DML operations, result set streaming, and Snowpark for Python-native data processing. The SQL API is a REST interface for submitting SQL statements without a JDBC/ODBC driver.

SurrealDB MCP server enabling AI agents to interact with SurrealDB — the multi-model database supporting relational, document, and graph data in a single engine — executing SurrealQL queries, managing tables and records, accessing graph traversal capabilities, and integrating SurrealDB's flexible data model into agent-driven application development and real-time data workflows.

Taiwan Stock Exchange (TWSE) MCP server enabling AI agents to access Taiwan stock market data — querying stock prices, company financial data, market indices (TAIEX), trading volumes, and Taiwan-listed company information from the official TWSE data source. Enables agents to research and analyze Taiwan equity markets.

Unreal Engine Motion Graphics MCP server enabling AI agents to interact with Unreal Engine's motion graphics and broadcast capabilities — controlling UE5 sequences, managing motion graphics templates, triggering animations, accessing real-time rendering output, and integrating AI-driven content automation into Unreal Engine broadcast and motion graphics production workflows.

Functional programming primitives for Ruby — Result (Success/Failure), Maybe (Some/None), Try, Task, and List monads. dry-monads enables railway-oriented programming in Ruby: chain operations where each step can succeed or fail, automatically propagating failures without nested if/else or exception handling. The `do` notation (Dry::Monads::Do) enables clean sequential composition with automatic failure propagation — like async/await but for Result types. Used extensively in Hanami and dry-transaction for composing service objects.

Docker-based MCP server enabling AI assistants to execute Kubernetes CLI tools (kubectl, helm, istioctl, argocd) in a secure containerized environment. Supports AWS EKS, Google GKE, and Azure AKS with non-root execution and strict command validation.

Safe, ergonomic single-assignment cells and lazy initialization for Rust. once_cell provides OnceCell<T> (single assignment), Lazy<T> (lazy initialization with a closure), and their thread-safe variants (sync::OnceCell, sync::Lazy). Solves the 'global constant that requires computation at startup' problem — global regex, compiled schemas, DB connections — that lazy_static and std::sync::Once handle more verbosely. Most of once_cell's API was stabilized in std (OnceLock, LazyLock) in Rust 1.70+.

Reads key-value pairs from .env files and sets them as environment variables — 12-factor app configuration pattern for Python. python-dotenv features: load_dotenv() to load .env into os.environ, dotenv_values() to load as dict without setting env vars, find_dotenv() to auto-locate .env up the directory tree, override= parameter to control whether existing env vars are overridden, dotenv_path parameter for explicit file path, Variable expansion (${OTHER_VAR}), comment support (#), multiline values with quotes, encoding parameter, and stream support for reading from file-like objects.

An MCP server that transforms codebases into searchable, hierarchical feature graphs by combining Tree-sitter AST parsing (43 languages), spectral clustering, and Obsidian-style wikilink navigation. Provides 11 MCP tools for structural analysis, semantic code search, blast radius tracing, static analysis, and safe code modification with shadow restore points.

Compile-time code generator for Java bean-to-bean mapping — the standard solution for DTO/entity mapping in Java applications. MapStruct generates type-safe, performant mapping implementations at compile time via annotation processing: define a @Mapper interface with method signatures, MapStruct generates the implementation. Eliminates reflection-based mappers (ModelMapper, Dozer) which have runtime overhead and no compile-time safety. Key features: automatic field mapping by name, custom mapping methods, decorator pattern, Spring/CDI injection, null-safety, and collection mapping.

Comprehensive weather data API providing current conditions, forecasts, historical data, and weather alerts for any location worldwide via REST endpoints.

Open-source swarm intelligence engine enabling self-governing AI agent collectives with queen/worker architecture, quorum voting, and local-first execution via MCP. Supports local models (Ollama), Claude Code CLI, Codex CLI, and paid APIs. Features include EVM wallet integration (USDC/USDT across multiple chains), semantic memory with vector embeddings, skill versioning, self-modification with audit trails, task scheduling (cron/webhook), and ERC-8004 on-chain identity.

The most widely-used testing framework for Scala — provides multiple test styles to match team preferences (FunSuite, FlatSpec, WordSpec, FeatureSpec, PropSpec) and rich matchers DSL. ScalaTest integrates with JUnit, Mockito, ScalaCheck (property-based testing), and all major Scala build tools (sbt, Maven, Gradle). Key features: flexible test style mixing, should/must/can matchers for readable assertions, async test support for Futures/IO, fixtures for shared setup/teardown, and Selenium integration for UI testing. The test style can be FunSpec (like RSpec) or FlatSpec (like Cucumber's flat syntax).

A security-focused runtime by Microsoft that executes WebAssembly Components via MCP, enabling AI agents to dynamically load and run sandboxed tools from an OCI registry without leaving the chat interface, using Wasmtime for browser-grade isolation.