Feed

`topease-customs-data` 是一个 MCP（Model Context Protocol）服务器，对接 [tradee.topease.net](https://tradee.topease.net/) 海关智能体贸易数据平台，支持按**进出口企业名称**、**产品关键字**、**HS编码**、**贸易时间**、**进出口类型**及**国家/市场/地区**进行多维度联合查询，帮助大模型快速获取真实的全球贸易情报数据。

OpenSERP MCP Server gives AI agents multi-engine web search (Google, Yandex, Bind, Baidu, Duckduckgo, Ecosia), SERP results, image search, mega-search, and URL extraction for LLM grounding, RAG, SEO, and automation.

LemmeBuyIt is a remote MCP server that gives AI agents access to real-time product search and pricing across 86 major US retailers — Walmart, Target, Home Depot, Best Buy, Macy's, Nordstrom, Zappos, Sephora, Tractor Supply, and 77 more. Approximately 262M products are indexed and refreshed continuously. Connect once with one-click OAuth (no API keys to copy) and the agent can: - Search any retailer's catalog by keyword, GTIN/UPC barcode, brand, or category - Look up a specific product by SKU and pull current price, stock status, reviews - Compare prices across retailers for the same item - Pull weekly price history (up to 2 years) for trend or deal analysis - Pull Amazon ASIN price + Buy Box history alongside the cross-retailer data The server exposes 7 read-only tools. No write access, no destructive operations. Free tier: search + product lookups. Paid tier: price-history tools.

Knowledge that gets sharper every time an agent uses it. A living knowledge base built by AI agents, for AI agents. Search community-verified solutions, share what you learn, turn dead ends into open problems, and converge on truth.

Bring your social media analytics, video intelligence, and competitor research into any MCP-compatible AI agent. SocialGPT is a hosted, remote MCP server - nothing to install, no API keys. Point your client at https://mcp.gpt.social/mcp, authenticate once with your SocialGPT account over OAuth 2.1 (PKCE), and your agent can: 1. Analyze your own content - list and search your TikTok, Instagram, and YouTube videos; pull deep per-video analysis (hooks, themes, transcripts, scene-by-scene breakdowns); and track views, engagement, and growth over time. 2. Research competitors - look up any public creator's profile and content DNA, browse their top and recent videos, find videos similar to yours, and benchmark against comparable creators. 3. Analyze new content on demand - drop in any TikTok/Instagram/YouTube URL or creator handle to scrape and analyze it into your library. 4. Run guided workflows - built-in prompts chain the tools into a full account audit, a head-to-head video benchmark, or a reverse-engineered playbook for any niche.

Extensive MCP tools so Claude Code, Cursor, and other MCP clients can create, edit, close, recap, and link your dev tasks while you code. Per-task open questions, attached resources, Jira linking, productivity recap.

Geo-based flight search MCP server. Find more flights between any two places on earth — connects directly to airlines for transparent, airline-sourced pricing with no hidden markups. Three tools: search_locations, get_city, search_flights. Supports one-way, return, and multi-city itineraries.

The official Massed Compute MCP server lets AI agents and developer tools check live GPU inventory, launch and tear down VM instances, and audit billing — all through natural-language tool calls. Connect over stdio (npx/uvx) or a hosted Streamable HTTP endpoint with a Bearer API key.

Persistent memory for AI agents. Give your coding assistant organizational context that compounds — search memories with hybrid retrieval, store anything with auto entity extraction, and query a living knowledge graph that gets smarter over time.

Access 300+ travel APIs through a single MCP. Look up airport data and codes, search flights and airlines, find hotels via Booking.com, check currency exchange rates, and more. Pay-per-call in USDC via x402 on Base — no API keys, no subscriptions.

Access 2,300+ e-commerce and retail APIs through a single MCP. Look up barcodes and UPC codes, search Walmart and eBay products, parse product pages, and more. Pay-per-call in USDC via x402 on Base — no API keys, no subscriptions.

Access 8,000+ web research and data extraction APIs through a single MCP. Search the web, scrape content, query Reddit, pull Google results, extract structured data, and more. Pay-per-call in USDC via x402 on Base — no API keys, no subscriptions.

Access 1,900+ blockchain and DeFi APIs through a single MCP. Estimate gas fees, analyze wallets, score on-chain volume, label addresses, and more. Pay-per-call in USDC via x402 on Base — no API keys, no subscriptions.

Pinako AI Bridge is the MCP server that connects Pinako, a Chromium extension for tab, window, library, and bookmark management, to local AI clients (Claude Desktop, Claude Code, Cursor, Windsurf, Cline, Roo Code, Continue.dev, and any HTTP MCP client). Because the bridge runs inside your desktop AI client alongside its other connections (Google Drive, Notion, GitHub, your filesystem, and more), the same assistant can move information between Pinako and everything else in one conversation. It's also a full agentic surface for Pinako itself: across 44 read and write tools, your AI can organize, tag, dedupe, summarize, and batch-edit your tabs, libraries, notes, and bookmarks, all undoable.

A brutally honest senior-engineer code reviewer as an MCP server for Claude Code, Cursor & Windsurf: 0–10 scores, real issues, usable fixes. 8 tools, free tier, no account needed.

Agent-callable Parallax services — list_services, get_pricing, start_project, book_call

A Model Context Protocol server that gives AI Clients real-time cybersecurity reconnaissance capabilities — WHOIS, DNS enumeration, port scanning, SSL inspection, CVE lookup, IP reputation, ASN lookup and more.

One persistent tmux-backed terminal for AI agents. ssh, docker exec, and REPLs nest inside a single authenticated session (no re-auth per command). Token-reduced reads, 4-layer completion detection, and a human can co-drive via the shared tmux socket. Install: npx -y aiterm-mcp

灵计一动:把灵感拆成可勾选的计划,AI 写代码时经 MCP 自动打钩,做完一键归档成文章。免费。 Lingji turns a one-line idea into a checkable plan outline. Connect your AI coding tool (Claude Code, Cursor, Trae) via MCP and it auto-checks tasks as it writes code — progress syncs to the web in real time, and editing the plan never interrupts the running AI. When done, archive the finished project into a publishable article draft. Free to use.

Inspect and manage IONOS CLOUD infrastructure via MCP

The social media publishing API that fails loudly instead of silently. Open source, idempotent by default, with preflight validation across 8 platforms: Bluesky, X, LinkedIn, Pinterest, Instagram, Facebook, Threads, and TikTok.

Build and send email, SMS, push, and web-push from your AI coding agent

The social media publishing API that fails loudly instead of silently. Open source, idempotent by default, with preflight validation across 8 platforms: Bluesky, X, LinkedIn, Pinterest, Instagram, Facebook, Threads, and TikTok.

Private agent-to-agent people matching. Contact only on a real match.

Structural graph memory for AI coding assistants. repo-graph maps your codebase entities, relationships, and feature flows so the model navigates to the right files instead of reading everything first. Tree-sitter, 20+ languages, frontend to backend and more, any MCP client.

Create and manage scheduled, guarded AI agent jobs with built-in quality control and 900+ connectors. Tell your agent what you want done, and Golemry turns it into an automation that runs on a schedule in the cloud. An overseer checks every result and escalates to you when something looks off, so silent failures don't slip through.

Most SEO and analytics MCP servers hand the raw rows back to the model: thousands of query/page records streamed into the context window for the LLM to crunch. Language models are not built for arithmetic over large tables. They are slow at it, they burn tokens doing it, and they make mistakes (dropped rows, miscounted sums, hallucinated totals) that are hard to catch. GSC Wizard does the opposite. Every analysis (CTR curves, decay detection, cannibalization, opportunity scoring, path breakdowns, ranking changes, full SEO reports) is computed in server-side Python and SQL, against the data warehouse, before anything reaches the model. The tool returns the finished result, not the raw input.