Best Token AI Skills & MCP Servers

Open-source CLI to minimize LLM token usage via smart code context extraction

Solana Token Safety API & MCP Server — rug pull, honeypot, and safety analysis for AI agents

ThryxProtocol v3.1.1 MCP Server — gasless AI agent launchpad + trading platform on Base. 21 tools wrapping the thryx.fun HTTP API. Server-managed wallets, paymaster gas, free LLM-driven autotrader. No private keys, no signing, no API bills. Auto-registers

Time Estimation MCP Server — 24 structured tools for PERT, COCOMO II, Monte Carlo, sprint forecasting, token-to-time mapping, cost estimation, schedule risk, and model comparison. Self-improving engine with estimation feedback loop. MCP + CLI + REST API.

MCP server for database connectivity. Multi-DB (PostgreSQL, MySQL, SQLite), connection management, schema introspection, query execution with token-optimized responses, rollback, history, elicitation for destructive ops, and prepared statements. Zero conf

Token-efficient browser MCP server — structured web pages for AI agents, not raw accessibility dumps

dwic — the design auditor for Claude Code. MCP server that audits your design tokens + markup for WCAG contrast failures, mandated-accent drift, and structural gaps, then remembers findings across sessions.

AI-to-AI token exchange + BSV blockchain. 20 MCP tools: reasoning, trading, NFTs, oracles, models, notarization. Trade tokens for on-chain assets.

GARL Protocol MCP Server — cryptographic verification for AI agent actions. 28 named tools: submit signed Action Receipts (v0.1), query multi-dimensional Trust Vector, issue/verify/revoke capability tokens (JWT-shaped + ECDSA-secp256k1), Capability Gate p

Design-system memory for coding agents — give Claude Code, Cursor, and Codex memory of your tokens, components, and Figma so they edit your app without breaking the design.

MCP server & CLI: codebase index + cache for Claude Code, Cursor, OpenAI Codex, Google Antigravity/Gemini. Model Context Protocol tools (codetree_read, codetree_search, …), hooks on Claude, rules on other IDEs. Token savings, SQLite index, symbol search,

AI agent OS for vibe coders — MCP server for Claude, Cursor, Windsurf & Lovable. Security audits, SEO, performance, design tokens, copywriting and 76+ dev tools in one CLI.

MCP server for MoySklad (Moy Sklad) ERP/inventory API — products, stock, orders, counterparties, supplies, profit reports. 10 tools. Token bucket rate limiting (45 req/3s). Dual auth: Bearer token or Basic login/password.

Zero-config codebase context for AI agents. One command generates a ~250-token navigation map and auto-configures Claude Code, Cursor, and Aider.

SSH MCP Server — Let AI manage remote servers via SSH with zero-token SFTP file transfer, proxy support, and multi-server management using Model Context Protocol

Programmatic MCP client library for Node.js - connect, discover, and call tools on Model Context Protocol servers.

MCP server that translates design tokens between platforms — Tailwind, Figma, CSS to Material 3, SwiftUI, and CSS Variables

jwt-lab – A fast, secure, beautiful JWT CLI tool and MCP server for developers & AI agents. Encode, decode, verify, inspect, audit, and generate keys for JSON Web Tokens.

TweetSave - Twitter/X analysis without the token waste. Fetch tweets, download media. Works with Cursor, Claude, VS Code, Gemini & more. Zero config → tweetsave.org

Model Context Protocol (MCP) server designed for LLMs to interact with Obsidian vaults. Provides secure, token-aware tools for seamless knowledge base management through a standardized interface.

Versioned memory for AI agents — structured keypaths, conflict detection, and 80% fewer tokens than alternatives.

Extract design tokens (colors, typography, spacing, border radius, shadows) from any public website. Generates JSON and CSS custom properties for local projects. Available as an AI agent skill (Claude, Cursor, Codex) and standalone CLI.

MCP server that analyzes your MCP setup: token costs, bloated schemas, duplicate tools, and optimization tips

MCP proxy that compresses prose fields (tool descriptions, etc.) using caveman rules. Same accuracy, fewer context tokens.