ツール

Gives agents controlled access to MySQL: list databases and tables, inspect schemas and sample data, and run SQL, with multi-database mode, SSL and SSH tunnels. Configure your own database user, ideally with least privilege.

Pulls new papers from Hugging Face Daily Papers and arXiv, scores them by research area into a recommendation page, and writes deep-read Obsidian notes for key ones; can also batch-read Zotero collections. Needs Python 3, Obsidian and poppler; copy the whole skills folder per the README.

LangChain's official skill collection: framework selection, quickstarts, middleware, persistence, human-in-the-loop approval, RAG and eval design across LangChain, LangGraph and Deep Agents, in Python and TypeScript. Also installable as a Claude Code plugin. Still in early development.

Connects agents to live Jupyter notebooks: read, write, insert and execute cells and read multimodal outputs like charts, for interactive data analysis. Bring your own JupyterLab or JupyterHub; open source, maintained by Datalayer.

MathWorks' official toolkit for reading, building, modifying and testing Simulink models via the MATLAB MCP server, with skills for control design, Simscape, code generation and compliance checks. Needs licensed MATLAB R2023a+ and Simulink; run the official installer in MATLAB first.

PortSwigger's official Burp Suite extension that lets agents read proxy history, send HTTP requests and create Repeater tabs; with Burp Pro it can also read scan results and use Collaborator. Install from the BApp Store; listens on 127.0.0.1:9876 by default.

The standard machine translation scorer: computes BLEU, chrF and TER with significance tests and downloads public test sets such as WMT, to compare engines or prompts. Needs Python 3.9+ and reference translations; automatic metrics are a rough screen, so final copy still needs human review.

Microsoft's official local MCP server for creating and editing Power BI semantic model tables, measures and relationships in natural language, running and validating DAX queries, and bulk renaming and translation. Connects to Power BI Desktop, Fabric or PBIP files; in public preview.

MathWorks' official toolkit: set up the MCP server via the MATLAB installer so agents can run code, run tests and do static checks; domain skill sets for signal processing, controls, statistics and machine learning install as Claude Code plugins. Needs MATLAB R2021a+.

Pre-submission review for CS conference papers (CVPR, ACL, NeurIPS and more): polish, translate or condense on request, or let AI reviewers raise issues and a jury rule on them. Only wording fixes are drafted, with your approval; missing experiments go back to you, never invented. Needs Node.js.

Connects agents to MongoDB and Atlas: query, aggregate, inspect collection schemas and indexes, and manage Atlas clusters, projects and database users. Maintained by MongoDB; needs a connection string or an Atlas service account; supports read-only mode.

Turns ComfyUI workflows into MCP tools with no code, covering generation and conversion across text, image, speech and video; works with local ComfyUI or RunningHub in the cloud. Start the server with uvx pixelle and configure an LLM first.

Skills Sentry's engineering team uses itself: code review, bug finding, security review, writing commits and PRs, and iterating on a PR until CI passes. Maintained by Sentry; some skills follow its internal conventions.

Microsoft's official Microsoft 365 MCP: query your email, meetings, OneDrive and SharePoint files, Teams messages and colleagues in plain language, e.g. tomorrow's meetings or someone's budget emails. Needs a Microsoft 365 Copilot license, tenant admin consent and Node.js 18+; public preview.

ComfyUI's official CLI: install and launch ComfyUI in one step, manage custom nodes and model downloads, and run image and video workflows, with --json output on every command for agents. Can also submit jobs to paid Comfy Cloud.

HR operations plugin bundling four skills: recruiting (job descriptions, interview scorecards, funnel analysis), HRBP (talent planning, performance, employee relations), people analytics (attrition prediction, pay equity) and ops, with helper Python tools. Scripts need Python 3.

A large skill library for non-engineering teams: 372 skills covering project management, marketing, executive advisory, legal (NDA triage, contract review, privacy, DPIAs), SOC 2 and GDPR compliance, HR and finance. Add it from the Claude desktop plugin page. Have legal output reviewed.

Graph database MCP server from Neo4j Labs: reads the graph schema and runs read and write Cypher queries, with read-only mode, query timeouts and response token limits. Needs a Neo4j instance with the APOC plugin enabled.

250+ human-reviewed legal skills from practicing lawyers, in-house counsel and academics, mostly US and EU law: compliance, GDPR, legal research, litigation, contract review and drafting, the EU AI Act. Install all or pick one; output is a draft for a lawyer to review.

Runs SQL queries against ClickHouse and browses databases and table schemas, read-only by default; can also enable embedded chDB to query data directly in files and URLs. Maintained by ClickHouse.

Jina AI's official remote MCP: converts web pages and PDFs to Markdown, searches the web, arXiv, SSRN and images, and offers screenshots, reranking and semantic dedup. Reading tools work without a key (rate-limited); search and rerank need a Jina API key.

Two parts: an API server on Kali Linux and an MCP bridge, letting agents run nmap, sqlmap, Metasploit, Hydra and other tools, or arbitrary terminal commands, for authorized penetration testing and CTFs. Available in Kali's official repositories.

Unbabel's machine-translation quality metric: comet-score rates translations per sentence and overall, XCOMET marks error spans and severity, and CometKiwi works without a reference. Good for spot-checking MT. Needs Python 3.8+; models download from Hugging Face, some need a license; GPU helps.

A research plugin for Claude Code: commands like /omp:survey, /omp:experiment and /omp:write chain literature review, topic selection, experiments, writing and self-review, with 5 specialized agents and 34 research skills that remember project progress across sessions. Experiments lean toward ML.