ツール

MiniMax's official CLI, built for agents: chat with MiniMax models from the terminal and generate images, video and speech, plus speech transcription, image understanding and web search. Works with the international platform. Requires a MiniMax Token Plan subscription.

A CLI and Python library by Simon Willison: imports JSON and CSV straight into SQLite with automatic table creation, runs in-memory SQL queries over files, and supports full-text search and table transformations.

Turns Figma designs into production code in your project, using the design screenshot as the visual target and reusing existing components, design tokens and Code Connect mappings first. Maintained by Figma; needs the Figma MCP server (free plans get only 6 calls a month).

Figma's official skill collection: implement designs as code, map components with Code Connect, generate design-system libraries from code, and build screens and FigJam diagrams on the canvas. Needs the Figma MCP server; canvas writes are in beta and currently free.

Figma's official plugin: connects the remote Figma MCP server and bundles skills for design-to-code, Code Connect, and generating designs and libraries, letting agents read design context and write to the canvas. Needs a Figma account; Starter plans and View/Collab seats get 6 calls a month.

Self-hostable long-term memory for agents, offering MCP, a REST API and a web dashboard: semantic search, knowledge graph and automatic memory consolidation. Embeddings are computed locally; choose SQLite, Cloudflare or Milvus as the backend.

AI localization: deploy the server to your own Vercel account (Postgres and AI Gateway included; enable deployment protection afterwards), then run npx languine@selfhosted translate to translate locale files with LLMs. A GitHub Action can open PRs. Needs a Vercel account and Node.js.

Recommends daily papers from arXiv and Semantic Scholar by research interest, reads them and writes illustrated Obsidian notes and a paper graph; can also search venues like CVPR and NeurIPS. Needs Python 3, Obsidian, a vault path and an interests config; the web app needs an Anthropic API key.

MathWorks' official server that lets agents start your local MATLAB, write and run scripts and unit tests, and check code style and potential errors, for computing, simulation and data analysis. Needs a licensed MATLAB R2021a+ and the downloaded server binary.

Built by Callstack from The Ultimate Guide to React Native Optimization: diagnoses dropped frames, slow startup, bundle size, memory leaks, unnecessary re-renders and janky animations, across both JS and native iOS/Android.

Lets agents drive WinDbg's CDB/KD to analyze crash dumps, debug user-mode processes remotely and debug kernels, and triage many dumps in a batch. Windows only; requires WinDbg or Debugging Tools for Windows.

Drives SolidWorks from prompts: part modeling, assemblies, drawings and STEP/DXF/PDF export, with multi-view previews for self-checks. A capability list marks what's verified; sheet metal, weldments and FEA are pilots needing engineer review. Needs Windows, Python 3.10+ and licensed SolidWorks.

Figma's official CLI that links Figma design-system components to real components in your code, so Dev Mode and the Figma MCP show production snippets instead of generated examples. Needs Node.js 18+, a Figma Organization or Enterprise plan, a full Design or Dev Mode seat and an access token.

Uses the Qdrant vector database as a semantic memory layer for agents: store information and retrieve it by meaning, connecting to a remote instance or a local path, with FastEmbed generating embeddings by default. Maintained by Qdrant.

KiCad design review skills: parse schematics, PCBs and Gerbers, trace the power tree, catch common mistakes, run SPICE simulation and EMC pre-checks, and tidy the BOM. Needs Python 3.10+; simulation needs ngspice; DigiKey needs API credentials. EMC pre-checks don't replace certification.

Condenses verbose xcodebuild and swift test output into concise build errors, warnings and test results, so agents read less log noise. Can generate JUnit reports and works with CI such as GitHub Actions; runs on macOS and Linux.

Connects LLMs to ROS and ROS 2 robots: publish and subscribe to topics, call services and actions, set parameters, and read sensor and robot state, without changing robot code. The robot side needs rosbridge running.

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.

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.

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.

Lets agents build Xcode projects and drive the iOS Simulator: 29 scripts cover building and testing, semantic navigation and taps via the accessibility tree, screenshot diffs and device lifecycle, with compact token-saving output. Needs macOS, Xcode and idb.

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.

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.