opennews-mcp-news-aggregation

# OpenNews MCP News Aggregation > Skill by [ara.so](https://ara.so) — MCP Skills collection. OpenNews MCP is a Model Context Protocol server providing real-time access to 84+ crypto and financial news sources across 6 categories (News, Listing, OnChain, Meme, Market, Prediction). Every article includes AI-powered impact scores (0-100), trading signals (long/short/neutral), and bilingual summaries. ## Installation ### Prerequisites 1. Get your API token from [https://6551.io/mcp](https://6551.io/mcp) 2. Set the token as an environment variable: ```bash export OPENNEWS_TOKEN="your-token-here" ``` ### Claude Desktop Add to `claude_desktop_config.json`: ```json { "mcpServers": { "opennews": { "command": "uv", "args": [ "--directory", "/path/to/opennews-mcp", "run", "opennews-mcp" ], "env": { "OPENNEWS_TOKEN": "your-token-here" } } } } ``` ### Using `claude mcp add` ```bash claude mcp add opennews \ -e OPENNEWS_TOKEN=your-token-here \ -- uv --directory /path/to/opennews-mcp run opennews-mcp ``` ### OpenClaw ```bash export OPENNEWS_TOKEN="your-token-here" cp -r openclaw-skill/opennews ~/.openclaw/skills/ ``` ## Configuration The server supports both environment variables and a `config.json` file in the project root. Environment variables take precedence. ### Environment Variables | Variable | Required | Default | Description | |----------|----------|---------|-------------| | `OPENNEWS_TOKEN` | **Yes** | - | API Bearer Token from 6551.io | | `OPENNEWS_API_BASE` | No | `https://ai.6551.io` | REST API base URL | | `OPENNEWS_WSS_URL` | No | `wss://ai.6551.io/open/news_wss` | WebSocket URL | | `OPENNEWS_MAX_ROWS` | No | `100` | Max results per request | ### config.json ```json { "api_base_url": "https://ai.6551.io", "wss_url": "wss://ai.6551.io/open/news_wss", "api_token": "your-token-here", "max_rows": 100 } ``` ## Data Sources Overview The platform aggregates 84+ sources across 6 engine types: - **news** (53 sources): Bloomberg, Reuters, Financial Times, CNBC, CoinDesk, Cointelegraph, The Block, Twitter/X, Telegram, and more - **listing** (9 sources): Binance, Coinbase, OKX, Bybit, Upbit, Robinhood, Hyperliquid listings - **onchain** (3 sources): Hyperliquid whale trades, large positions, KOL trades - **meme** (1 source): Twitter meme coin sentiment - **market** (6 sources): Price changes, funding rates, liquidations, OI changes - **prediction** (12 sources): AI correlation, smart money, whale positions, insider patterns ## Available Tools ### Discovery Tools #### `get_news_sources` Retrieves the complete engine tree with all categories and sources. ```python # Returns hierarchical structure of all 6 engine types and their sources # Each source includes metadata like description and type ``` **Response structure:** ```json { "news": { "Bloomberg": {"description": "Bloomberg — top-tier financial news"}, "CoinDesk": {"description": "CoinDesk — leading crypto media"} }, "listing": { "Binance": {"description": "Binance new token listings"} }, "onchain": { "Hyperliquid Whale Trade": {"description": "Hyperliquid whale trade alerts"} } } ``` #### `list_news_types` Returns a flat list of all available source codes for filtering. ```python # Returns array of all source codes across all engine types # Use these codes in other tool filters ``` **Response:** ```json ["Bloomberg", "Reuters", "CoinDesk", "Binance", "Hyperliquid Whale Trade", ...] ``` ### Search Tools #### `get_latest_news` Fetch the most recent articles across all sources. **Parameters:** - `limit` (optional, default: 20): Number of articles to return ```python # Get latest 50 articles {"limit": 50} ``` #### `search_news` Full-text keyword search across all sources. **Parameters:** - `keywords` (required): Search query string - `limit` (optional, default: 20): Number of results ```python # Search for SEC-related news {"keywords": "SEC regulation", "limit": 30} ``` #### `search_news_by_coin` Filter articles by specific cryptocurrencies. **Parameters:** - `coins` (required): Array of coin symbols (e.g., `["BTC", "ETH"]`) - `limit` (optional, default: 20): Number of results ```python # Get Bitcoin and Ethereum news {"coins": ["BTC", "ETH"], "limit": 40} ``` #### `get_news_by_source` Filter by specific source within an engine type. **Parameters:** - `engine_type` (required): Engine category (`news`, `listing`, `onchain`, `meme`, `market`, `prediction`) - `news_type` (required): Specific source code (e.g., `Bloomberg`, `Binance`) - `limit` (optional, default: 20): Number of results ```python # Get Bloomberg news only {"engine_type": "news", "news_type": "Bloomberg", "limit": 25} # Get Binance listings {"engine_type": "listing", "news_type": "Binance", "limit": 10} ``` #### `get_news_by_engine` Filter articles by engine category. **Parameters:** - `engine_type` (required): Engine category - `limit` (optional, default: 20): Number of results ```python # Get all on-chain events {"engine_type": "onchain", "limit": 30} # Get all prediction signals {"engine_type": "prediction", "limit": 50} ``` #### `search_news_advanced` Multi-filter search combining coins, keywords, and engine types. **Parameters:** - `coins` (optional): Array of coin symbols - `keywords` (optional): Search query - `engine_types` (optional): Array of engine categories - `limit` (optional, default: 20): Number of results ```python # Bitcoin news from Bloomberg and CoinDesk mentioning "ETF" { "coins": ["BTC"], "keywords": "ETF", "engine_types": ["news"], "limit": 30 } # Whale activity for SOL and ETH { "coins": ["SOL", "ETH"], "engine_types": ["onchain"], "limit": 20 } ``` ### AI-Powered Tools #### `get_high_score_news` Retrieve articles with high AI impact scores. **Parameters:** - `min_score` (optional, default: 80): Minimum impact score (0-100) - `limit` (optional, default: 20): Number of results ```python # Get articles with score >= 90 {"min_score": 90, "limit": 15} # Get articles with score >= 70 {"min_score": 70, "limit": 30} ``` #### `get_news_by_signal` Filter by AI-generated trading signals. **Parameters:** - `signal` (required): Trading signal type (`long`, `short`, `neutral`) - `limit` (optional, default: 20): Number of results ```python # Get bullish signals {"signal": "long", "limit": 25} # Get bearish signals {"signal": "short", "limit": 25} # Get neutral signals {"signal": "neutral", "limit": 20} ``` ### Real-Time Tools #### `subscribe_latest_news` Subscribe to WebSocket live feed with optional filters. **Parameters:** - `engine_types` (optional): Object mapping engine types to source codes - Key: Engine type (`news`, `listing`, `onchain`, etc.) - Value: Array of source codes (empty array = all sources in that engine) - `coins` (optional): Array of coin symbols - `has_coin` (optional, boolean): Only articles tagged with coins ```python # Subscribe to Bloomberg and CoinDesk for BTC and ETH { "engine_types": { "news": ["Bloomberg", "CoinDesk"] }, "coins": ["BTC", "ETH"], "has_coin": True } # Subscribe to all listing announcements { "engine_types": { "listing": [] } } # Subscribe to whale on-chain activity { "engine_types": { "onchain": [] } } ``` **WebSocket message format (incoming):** ```json { "jsonrpc": "2.0", "method": "news.update", "params": { "id": "article-id", "title": "Article title", "content": "Full article text", "engine_type": "news", "news_type": "Bloomberg", "coins": ["BTC"], "ai_score": 85, "ai_signal": "long", "summary_en": "English summary", "summary_zh": "中文摘要", "published_at": "2026-05-16T12:00:00Z" } } ``` ## Response Data Structure All news articles follow this structure: ```json { "id": "unique-article-id", "title": "Article headline", "content": "Full article text content", "engine_type": "news", "news_type": "Bloomberg", "coins": ["BTC", "ETH"], "ai_score": 85, "ai_signal": "long", "summary_en": "English AI-generated summary", "summary_zh": "中文AI生成摘要", "published_at": "2026-05-16T10:30:00Z", "url": "https://original-source.com/article", "metadata": { "author": "Jane Doe", "tags": ["regulation", "ETF"] } } ``` **Key fields:** - `ai_score` (0-100): AI impact score, higher = more market-moving - `ai_signal`: Trading direction (`long`, `short`, `neutral`) - `coins`: Array of related cryptocurrency symbols - `engine_type`: Category (news, listing, onchain, meme, market, prediction) - `news_type`: Specific source code ## Common Usage Patterns ### Pattern 1: Monitor High-Impact News ```python # Get latest high-impact news (score >= 80) result = get_high_score_news(min_score=80, limit=10) # Filter for bullish signals only bullish = get_news_by_signal(signal="long", limit=15) # Combine: high-impact bullish Bitcoin news