Browse the catalog

An AI debugging agent MCP server that enables autonomous plan-act-observe debugging workflows, allowing repository exploration, code inspection, human-approved edits, and test execution through structured MCP tools.

RTK Context Compressor public documentation modeled after the MiroFish docs pattern.

An MCP wrapper for https://sv443.net/jokeapi/v2/

yuexia\_test

Enhanced Attio CRM API skill with batch operations.

Provides read-only access to your HackerOne reports, programs, earnings, and scope data via the HackerOne Hacker API. Search and filter reports by keyword, program, severity, or state. Analyze hunting patterns including severity distribution, top programs, and weakness types. Check program scope and accepted weaknesses before reporting.

This code execution server enables AI assistants to run Python and bash code in isolated Docker containers or virtual machines, providing a secure sandboxed environment for code execution tasks. Built with FastMCP and Docker integration, it supports both ephemeral containers (created from images) and persistent named containers, with configurable timeouts and environment variables. The implementation includes a Kali Linux-based container with pre-installed security tools and scientific packages, making it particularly useful for cybersecurity analysis, data science workflows, and general-purpose code execution where isolation and security are paramount.

An MCP server built on top of Nasdanika (AI) capabilities

Metabase MCP Server provides a bridge between AI assistants and Metabase analytics platforms, enabling conversational access to business intelligence data. Developed by Hyeongjun Yu, this TypeScript implementation exposes six tools that allow assistants to list dashboards, cards, and databases, execute saved questions with parameters, extract dashboard components, and run custom SQL queries. The server supports both session-based and API key authentication methods, delivers structured JSON responses, and includes comprehensive error handling and logging capabilities, making it particularly valuable for teams that want to analyze their Metabase data through natural language interactions.

Connects AI agents to Steep workspace for querying metrics, targets, entities, and team data via natural language. All tools are read-only and the server is remote.

Model Context Protocol (MCP) server tool JSON schema validator

A from-scratch implementation of the ModelContextProtocol to deepen my understanding.

Exposes a FastAPI task management REST API as MCP tools, enabling an LLM client to list, create, get, complete, and delete tasks via natural language.

Startup ecosystem intelligence for VC deal sourcing gives your AI assistant instant access to 8 public data sources — patents, GitHub activity, job postings, ArXiv research, tech stacks, corporate registries, and SaaS competitive data — all fused into a single structured deal memo.

Estimates token costs and compresses prompts before LLM requests, enabling teams to optimize spending and stay within budget constraints. Provides accurate pre-execution cost predictions across multiple LLM providers with configurable compression strategies.

Used vehicle inspection scoring and live dealer B2B auction platform (Carsome style)

Open-Meteo MCP — weather forecast + historical reanalysis + sister APIs

GapBase provides access to a curated database of 474 validated business opportunity gaps sourced from Reddit, LinkedIn, and X, organized across seven industries including ecommerce, healthcare, legal, and accounting. Each gap includes a problem statement and suggested solution direction, searchable by keyword or industry. The server operates entirely offline with no network calls or data collection, and surfaces five weekly trending gaps from live social signals.

Edge-quantized multimodal spatial point locator (MiniCPM-V / Molmo style)

opencode can do all the things related to code.