Packages

Code Runner MCP server enabling AI agents to execute code in multiple programming languages — running Python, JavaScript, TypeScript, Bash, Go, Rust, and other language snippets, capturing stdout/stderr output, and integrating local code execution into agent-driven development, testing, and data analysis workflows.

Google's open-source AI agent for the terminal that integrates Gemini models for code analysis, generation, automation, and general tasks. Supports MCP server integration, multimodal input, Google Search grounding, and deep GitHub workflow automation.

Official Steampipe MCP server from Turbot enabling AI agents to query cloud infrastructure, security configurations, and compliance data using SQL through Steampipe — the open-source tool that maps cloud APIs (AWS, Azure, GCP, Kubernetes, GitHub, etc.) to queryable SQL tables. Agents can run SQL queries across 140+ cloud services through a single consistent interface.

Turso cloud SQLite — a libSQL fork of SQLite with HTTP API access, replication, and database-per-tenant model for deploying isolated SQLite databases at the edge.

WordPress MCP server enabling AI agents to interact with WordPress sites — creating, reading, updating, and deleting posts and pages, managing media, handling comments, accessing user and taxonomy data, and integrating WordPress content management into agent-driven content creation and publishing workflows.

An MCP server that enables AI assistants to automate macOS through AppleScript and JavaScript for Automation (JXA). Includes 200+ pre-built automation recipes and accessibility API integration for UI element control.

Datadog MCP server enabling AI agents to interact with Datadog's observability platform — querying metrics, logs, and traces; managing monitors and alerts; investigating incidents; running Datadog Query Language (DQL) queries; and integrating comprehensive observability data into agent-driven SRE and incident response workflows.

MCP server for Directus — the open-source headless CMS and data platform. From rijkvanzanten, the creator of Directus. Enables AI agents to interact with Directus collections, items, users, files, and settings through Directus's REST and GraphQL API — supporting AI-driven content management and data workflows on Directus deployments.

HashiCorp Vault is the industry-standard open-source secrets management platform. Its HTTP API is the sole interface for all Vault operations: authenticating workloads (token, AppRole, Kubernetes, AWS IAM, LDAP, OIDC, and 15+ other auth methods), reading and writing secrets (KV v1/v2, database credentials, cloud IAM keys, SSH certificates, PKI certificates), managing leases and TTLs, configuring policies, and administering Vault clusters. For agents, Vault is uniquely powerful because it generates dynamic, short-lived credentials on demand — an agent asks for a Postgres password, Vault creates a dedicated DB user with a 1-hour TTL, returns the credentials, and automatically revokes them when the lease expires. This eliminates long-lived static secrets from agent workflows entirely. AppRole is the canonical machine auth method: a RoleID (public, embedded in config) plus a SecretID (private, injected at runtime) produces a Vault token with specific policies. Kubernetes auth allows pods to authenticate using their ServiceAccount JWT without any pre-shared secrets. The API is entirely REST-over-HTTPS with JSON request/response bodies and uses the X-Vault-Token header for authenticated requests.

Official InfluxDB 3 MCP server enabling AI agents to query and write time-series data — running Flux and SQL queries against InfluxDB, writing metrics/events, managing buckets, and integrating with IoT and observability pipelines.

Official Atlassian MCP server providing agent-native tools for searching, creating, and updating Jira issues and projects — enabling agents to interact with Jira Cloud without constructing raw REST calls or managing JQL queries manually.

TypeScript MCP server template and example implementation — providing a well-structured starting point for building custom MCP servers in TypeScript, demonstrating best practices for tool definition, error handling, and MCP protocol implementation using the official TypeScript MCP SDK.

Nix flake packaging of MCP servers — providing reproducible, declarative Nix packages for popular MCP servers, enabling NixOS and Nix flake users to install and configure MCP servers through Nix's dependency management, ensuring reproducible builds and consistent MCP server environments across machines.

Provides a long-term memory layer for AI agents that stores, retrieves, and manages structured memories using vector embeddings and an optional knowledge graph, enabling persistent personalization across sessions.

Official Vercel MCP server that gives AI coding agents runtime diagnostics, development automation, and documentation access for Next.js applications — including real-time error detection, route inspection, automated Next.js 16 upgrades, and browser testing via Playwright.

PostgreSQL MCP server enabling AI agents to interact directly with PostgreSQL databases — executing SQL queries, managing tables and schemas, reading and writing data, and integrating PostgreSQL database access into agent-driven data workflows for any PostgreSQL-compatible database.

Puppeteer MCP server enabling AI agents to control headless Chrome for browser automation — navigating web pages, taking screenshots, filling forms, extracting page content, executing JavaScript, and integrating Puppeteer's browser automation capabilities into agent-driven web scraping, testing, and web interaction workflows.

Redis Stack extends Redis with JSON document storage (RedisJSON), full-text and vector search (RediSearch), time-series data (RedisTimeSeries), and probabilistic data structures. Builds on Redis's in-memory key-value store with sub-millisecond latency. Accessible via the Redis protocol (RESP) using standard Redis clients, plus a REST API via Redis Stack Server. Available as Redis OSS + Stack modules, or Redis Cloud.

Official Shortcut MCP server enabling AI agents to interact with Shortcut (formerly Clubhouse) project management — creating, reading, and updating stories and epics, managing workflows and iterations, searching for stories, accessing member and team data, and integrating Shortcut project management into agent-driven sprint planning and engineering workflow automation.

An MCP server that provides LLMs with up-to-date, version-specific documentation and code examples for programming libraries. Resolves library names to Context7 IDs, then retrieves current documentation and examples directly into the agent's context. Prevents hallucinated or outdated API references by pulling from live sources.