gavindiaz/opennews-mcp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

docs: update discovery and news tools to reflect financial market focus

Browse Source 
- Revised descriptions in discovery.py, free.py, news.py, and realtime.py to emphasize financial market news sources and categories.
- Enhanced docstrings to clarify the scope of news coverage, including crypto, U.S. equities, macro, and other relevant signals.
- Updated function descriptions to align with the new focus on market news and signals.
This commit is contained in:
beare
2026-06-15 16:22:17 +08:00
parent 111e64faad
commit 9c6b508c23
4 changed files with 54 additions and 35 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/opennews_mcp/tools/discovery.py
+6 -4
View File
@@ -1,4 +1,4 @@
"""Discovery tools — list available news sources and categories."""
"""Discovery tools — list available financial market news sources and categories."""
from mcp.server.fastmcp import Context
@@ -8,7 +8,7 @@ from opennews_mcp.config import require_token
@mcp.tool()
async def get_news_sources(ctx: Context) -> dict:
"""Get all available news source categories and their metadata.
"""Get all available market news source categories and their metadata.
This platform aggregates 84+ real-time data sources across 6 engine categories:
@@ -16,7 +16,9 @@ async def get_news_sources(ctx: Context) -> dict:
CoinDesk, Cointelegraph, The Block, Blockworks, Decrypt, DlNews, A16Z, TechCrunch,
Wired, Politico, Business Insider, Twitter/X, Telegram, Weibo, Truth Social,
U.S. Treasury, ECB, TASS, Handelsblatt, Welt, Ambrey, Morgan Stanley (MS NOW),
PR Newswire, Coinbase, and more.
PR Newswire, Coinbase, and more. Useful for crypto, U.S. equities, macro,
semiconductors, AI infrastructure, supply chains, commodities, rates, policy,
and market-moving social/news signals.
LISTING (9 sources): Binance, Coinbase, OKX, Bybit, Upbit, Bithumb, Robinhood,
Hyperliquid, Aster — new token listing announcements from major exchanges.
@@ -75,7 +77,7 @@ async def get_news_sources(ctx: Context) -> dict:
@mcp.tool()
async def list_news_types(ctx: Context) -> dict:
"""List all available news type codes for filtering.
"""List all available market news type codes for filtering.
Returns a flat list of news source codes that can be used with
the newsType parameter in search_news.
src/opennews_mcp/tools/free.py
+7 -3
View File
@@ -1,4 +1,4 @@
"""Free news tools — no token required, access via /open/free_* endpoints.
"""Free financial market news tools — no token required, access via /open/free_* endpoints.
These tools provide basic news access without authentication.
For full features (84+ sources, AI analysis, real-time WebSocket), get a free token at https://6551.io/mcp.
@@ -11,7 +11,7 @@ from opennews_mcp.app import mcp
@mcp.tool()
async def get_news_categories(ctx: Context) -> dict:
"""Get all available news categories and subcategories.
"""Get all available free market news categories and subcategories.
Returns a list of categories, each containing subcategories,
for use with the get_hot_news tool.
@@ -32,7 +32,11 @@ async def get_hot_news(
ctx: Context,
subcategory: str = "",
) -> dict:
"""Get hot news and tweets by category.
"""Get hot market news and tweets by category.
Use this free endpoint for curated real-time financial market news, including
crypto, equities, macro, policy, commodities, and market-moving social signals
when the authenticated 84+ source search is unavailable.
Args:
category: Category key (required). Use get_news_categories to list available keys.
src/opennews_mcp/tools/news.py
+34 -23
View File
@@ -1,8 +1,9 @@
"""News content tools — search and retrieve news from 84+ sources via REST API.
"""News content tools — search and retrieve real-time financial market news via REST API.
Covers 6 engine categories: news (53 premium media & social sources), listing (9 exchanges),
onchain (whale & KOL trades), meme (social sentiment), market (6 anomaly signals),
prediction (12 AI prediction signals).
Covers crypto, U.S. equities, macro, semiconductors, AI infrastructure, supply chains,
commodities, rates, policy, and market-moving social/news signals across 6 engine categories:
news (53 premium media & social sources), listing (9 exchanges), onchain (whale & KOL trades),
meme (social sentiment), market (6 anomaly signals), and prediction (12 AI prediction signals).
Uses POST /open/news_search as the primary data source.
"""
@@ -14,10 +15,12 @@ from opennews_mcp.config import clamp_limit, make_serializable, MAX_ROWS, requir
@mcp.tool()
async def get_latest_news(ctx: Context, limit: int = 10) -> dict:
"""Get the most recent crypto news articles, newest first.
"""Get the most recent market-moving news and signals, newest first.
Returns news from 84+ sources across all 6 categories (news, listing, onchain,
meme, market, prediction) with title text, source, link, related coins, AI rating, and tags.
meme, market, prediction) with title text, source, link, related assets, AI rating, and tags.
Coverage includes crypto, U.S. equities, macro, semiconductors, AI infrastructure,
supply chains, commodities, rates, policy, and social/news signals.
Args:
limit: Maximum number of articles to return (default 10, max 100).
@@ -39,13 +42,15 @@ async def get_latest_news(ctx: Context, limit: int = 10) -> dict:
@mcp.tool()
async def search_news(keyword: str, ctx: Context, limit: int = 10) -> dict:
"""Search crypto news by keyword in text content.
"""Search real-time financial market news by keyword in text content.
Searches across all 84+ sources including Bloomberg, Reuters, CoinDesk,
Searches across all 84+ sources for crypto, U.S. equities, macro, semiconductors,
AI infrastructure, supply chains, commodities, rates, policy, and market-moving
social/news signals. Sources include Bloomberg, Reuters, FT, CNBC, CoinDesk,
Twitter/X, on-chain alerts, exchange listings, market signals, and AI predictions.
Args:
keyword: Search term (e.g. "bitcoin", "SEC", "ETF").
keyword: Search term (e.g. "bitcoin", "NVDA", "FOMC", "tariffs", "oil", "AI chips").
limit: Maximum results (default 10, max 100).
"""
if (err := require_token()):
@@ -65,7 +70,7 @@ async def search_news(keyword: str, ctx: Context, limit: int = 10) -> dict:
@mcp.tool()
async def search_news_by_coin(coin: str, ctx: Context, limit: int = 10) -> dict:
"""Search news related to a specific cryptocurrency coin/token.
"""Search news and market signals related to a specific digital asset coin/token.
Finds all mentions across 84+ sources: media coverage, exchange listings,
whale trades, meme sentiment, market anomalies, and AI predictions for the given coin.
@@ -91,7 +96,7 @@ async def search_news_by_coin(coin: str, ctx: Context, limit: int = 10) -> dict:
@mcp.tool()
async def get_news_by_source(engine_type: str, news_type: str, ctx: Context, limit: int = 10) -> dict:
"""Get news articles from a specific source.
"""Get market news or signal items from a specific source/category.
Use get_news_sources first to see available engine types and news type codes.
@@ -127,10 +132,12 @@ async def get_news_by_source(engine_type: str, news_type: str, ctx: Context, lim
@mcp.tool()
async def get_news_by_engine(engine_type: str, ctx: Context, limit: int = 10) -> dict:
"""Get news articles filtered by engine type.
"""Get market news or signal items filtered by engine type.
Engine types: "news", "listing", "onchain", "meme", "market", "prediction".
- "news": 53 sources — Bloomberg, Reuters, FT, CNBC, CNN, BBC, CoinDesk, Twitter/X, etc.
- "news": 53 sources — Bloomberg, Reuters, FT, CNBC, CNN, BBC, CoinDesk, Twitter/X, etc.,
covering crypto, U.S. equities, macro, semiconductors, AI infrastructure,
supply chains, commodities, rates, policy, and social/news signals.
- "listing": 9 exchanges — Binance, Coinbase, OKX, Bybit, Upbit, Bithumb, Robinhood, etc.
- "onchain": Whale trades & KOL activity on Hyperliquid.
- "meme": Meme coin social sentiment from Twitter.
@@ -166,16 +173,18 @@ async def search_news_advanced(
min_score: int = 0,
limit: int = 10,
) -> dict:
"""Advanced news search with multiple filters.
"""Advanced financial market news search with multiple filters.
Combines coin, keyword, engine type, and source filters for precise queries
across the full 84+ source catalog.
Combines keyword, digital asset coin, engine type, source, and score filters for precise
queries across the full 84+ source catalog. Use it for crypto, U.S. equities, macro,
semiconductors, AI infrastructure, supply chains, commodities, rates, policy, and
market-moving social/news signals.
Args:
coins: Comma-separated coin symbols (e.g. "BTC,ETH").
keyword: Optional search keyword.
keyword: Optional search keyword (e.g. "NVDA", "FOMC", "tariffs", "oil", "AI chips").
engine_types: Engine type filter in format "type1:cat1,cat2;type2:cat3" (e.g. "news:Bloomberg,Reuters;listing:;prediction:").
has_coin: If true, only return news that have associated coins.
has_coin: If true, only return items that have associated digital asset coins.
min_score: Minimum AI score threshold (default 0, range 0-100).
limit: Maximum results (default 10, max 100).
"""
@@ -215,10 +224,12 @@ async def search_news_advanced(
@mcp.tool()
async def get_high_score_news(ctx: Context, min_score: int = 70, limit: int = 10) -> dict:
"""Get highly-rated news articles (by AI score), sorted by score descending.
"""Get highly-rated market news and signals by AI score, sorted by score descending.
AI scores range 0-100 and reflect potential market impact.
All articles from 84+ sources are AI-analyzed with score, grade, signal, and summary.
AI scores range 0-100 and reflect potential market impact across crypto,
U.S. equities, macro, semiconductors, AI infrastructure, supply chains,
commodities, rates, policy, and social/news signals. All items from 84+
sources are AI-analyzed with score, grade, signal, and summary.
Args:
min_score: Minimum score threshold (default 70).
@@ -242,9 +253,9 @@ async def get_high_score_news(ctx: Context, min_score: int = 70, limit: int = 10
@mcp.tool()
async def get_news_by_signal(signal: str, ctx: Context, limit: int = 10) -> dict:
"""Get news filtered by trading signal type.
"""Get market news filtered by AI trading signal type.
Each article from 84+ sources is AI-analyzed for trading direction.
Each item from 84+ sources is AI-analyzed for directional market impact.
Args:
signal: The signal type: "long" (bullish), "short" (bearish), or "neutral".
src/opennews_mcp/tools/realtime.py
+7 -5
View File
@@ -1,4 +1,4 @@
"""Real-time news tools — WebSocket subscription for live updates from 84+ sources."""
"""Real-time market news tools — WebSocket subscription for live updates from 84+ sources."""
from mcp.server.fastmcp import Context
@@ -15,22 +15,24 @@ async def subscribe_latest_news(
engine_types: str = "",
has_coin: bool = False,
) -> dict:
"""Subscribe to real-time news updates via WebSocket.
"""Subscribe to real-time financial market news updates via WebSocket.
Connects to the WebSocket feed, subscribes to news with optional filters,
and collects incoming messages for the specified duration.
Streams live data from 84+ sources across 6 categories:
news (Bloomberg, Reuters, etc.), listing (Binance, Coinbase, etc.),
news (Bloomberg, Reuters, FT, CNBC, Twitter/X, etc.), listing (Binance, Coinbase, etc.),
onchain (whale trades), meme (Twitter sentiment), market (price/funding/liquidation alerts),
prediction (AI prediction signals — smart money, correlations, whale positions, etc.).
Coverage includes crypto, U.S. equities, macro, semiconductors, AI infrastructure,
supply chains, commodities, rates, policy, and market-moving social/news signals.
Args:
wait_seconds: How long to listen for news (default 10, max 30 seconds).
max_items: Maximum news items to collect (default 5, max 20).
coins: Comma-separated coin symbols to filter (e.g. "BTC,ETH").
coins: Comma-separated digital asset coin symbols to filter (e.g. "BTC,ETH").
engine_types: Engine type filter in format "type1:cat1,cat2;type2:cat3".
has_coin: If true, only receive news that have associated coins.
has_coin: If true, only receive items that have associated digital asset coins.
"""
if (err := require_token()):
return err