Ready-to-use agent config files for Codex, OpenCode, and Claude Code — drop in and customize.
6 MCP servers — puppeteer: Browser automation for visual testing, screenshots, accessibility audits, and UI verification.; figma: Read Figma designs, inspect component properties, extract design tokens and spacing values.; storybook: Browse and interact with Storybook component library. Inspect stories, props, and variants.; filesystem: Read and write component files, styles, tests, and configuration.; github: Manage PRs, review UI changes, and track frontend-related issues. (+1 more)
To connect Cursor to the Home Depot MCP server: 1. **In Cursor Settings:** 2. **Or create a config file:** "args": ["dist/index.js"], "cwd": "/Users/makaminski1337/Developer/depot-mcp"
This guide will walk you through setting up a multi-agent MCP server that provides specialized AI agents (planner, reviewer, architect) to work with Claude Code.
4 MCP servers — defi-mcp: DeFi and crypto data — token prices, gas fees, swap quotes, wallet balances, and trending tokens across 9 chains.; filesystem: Read and write strategy files, portfolio configs, and analysis outputs.; fetch: Fetch blockchain explorer data, protocol documentation, and governance proposals.; memory: Track portfolio positions, price alerts, and research findings across sessions.
6 MCP servers — jupyter: Create, edit, and execute Jupyter notebook cells. Manage kernels and inspect outputs.; sqlite: Query SQLite databases for local data exploration, schema inspection, and ad-hoc analytics.; postgres: Query the data warehouse for production analytics, reporting, and data validation.; filesystem: Read CSV, Parquet, and JSON data files. Write processed outputs and reports.; fetch: Fetch datasets from public APIs, download documentation, and access data catalogs. (+1 more)
7 MCP servers — aws: Manage AWS resources: EC2, S3, Lambda, ECS, RDS, CloudFormation, and IAM.; docker: Build, run, and manage Docker containers, images, volumes, and networks.; github: Manage repositories, Actions workflows, releases, and infrastructure-as-code PRs.; terraform: Plan, apply, and inspect Terraform state. Browse provider documentation and module registry.; kubectl: 270+ Kubernetes tools for cluster management, deployments, troubleshooting, and monitoring. (+2 more)
7 MCP servers — filesystem: Read, write, and manage project files and directories.; github: Manage repositories, issues, PRs, branches, and releases on GitHub.; postgres: Query PostgreSQL databases, inspect schemas, and run SQL statements.; redis: Interact with Redis for cache inspection, key management, and pub/sub.; puppeteer: Browser automation for testing UI, taking screenshots, and verifying rendered output. (+2 more)
6 MCP servers — kubectl: 270+ Kubernetes management tools: pods, deployments, services, helm, networking, RBAC, and cluster operations.; kubectl-app: 8 interactive UI dashboards for Kubernetes: pods, logs, deployments, helm, cluster, cost, events, network.; docker: Manage Docker containers, images, volumes, and networks for local development and testing.; github: Manage Kubernetes manifests, Helm charts, and CI/CD workflows in GitHub repositories.; filesystem: Read and write Kubernetes YAML mani
14 MCP servers — filesystem: Read, write, and manage files and directories. Replace /path/to/project with your actual project root.; github: Create issues, PRs, read repositories, manage branches and releases on GitHub.; postgres: Query PostgreSQL databases, inspect schemas, run read-only or read-write SQL.; redis: Interact with Redis: get/set keys, inspect data structures, manage cache.; docker: Manage Docker containers, images, volumes, and networks. (+9 more)
5 MCP servers — bgpt: Search scientific papers with full-text experimental data via BGPT. 50 free searches, no API key needed.; fetch: Fetch web pages, documentation, and research resources as markdown.; memory: Persistent knowledge graph for tracking research findings, citations, and insights across sessions.; filesystem: Read and write research notes, papers, and data files. Replace the placeholder with your project root.; brave-search: Web search for supplementary sources, preprints, and grey
Production-ready Claude Code settings.json with recommended configuration for development workflows.
Cursor IDE rules and conventions for AI-assisted development.
Document the mapping in your config file (no code yet). For per-session provider selection, store the chosen provider/model with the Session record.
<test_user> email: basic@testing.com password: BasicTest123$ role: user plan: basic id: 22222222-2222-2222-2222-222222222222 DO NOT ATTEMPT TO SOLVE THE CLOUDFLARE TURNSTILE </test_user> <application_pages> Main pages in our application:
Curated MCP server configuration for Claude Code with recommended tools and integrations.
2 MCP servers — n8n; pipedream
Cursor IDE rules and conventions for AI-assisted development.
1 MCP server — buywhere: Agent-native product catalog API for Southeast Asia commerce. Search 1.5M+ products across Shopee, Lazada, Amazon SG, and 50+ merchants. Real-time price comparison, deal discovery, and multi-currency support (SGD, USD, MYR, IDR, THB, PHP, VND). Tools: search_products, get_product, compare_prices, get_deals, browse_categories, get_category_products.
4 MCP servers — helium: Real-time news with bias scoring across 5,000+ sources, live stock/ETF/crypto data with AI bull/bear cases and price forecasts, ML options pricing with probability ITM, balanced news synthesis, and meme search. 9 tools, free tier (50 queries), no API key needed.; chart-library: Historical chart pattern search engine. Search 24M+ pre-computed embeddings across 15K+ symbols and 10 years of data. 19 tools for pattern matching, forward returns, regime analysis, anomaly detect
1 MCP server — llm-prices: Look up and compare LLM API costs across 167 models from 23 providers. Ask Claude 'what's the cheapest model for 10k input + 2k output tokens?' or 'compare gpt-4o vs claude-sonnet-4-6 for my workload'. No API key required — pricing data is bundled.