# wpnews.pro > AI-curated news aggregator covering artificial intelligence, developer tools, > cloud infrastructure, and technology policy. Updated every 30 minutes from > curated RSS feeds. All articles are full-text extracted, summarised by DeepSeek, > and available in JSON, Markdown, plain text, MCP, and NDJSON formats. ## MCP (Model Context Protocol — recommended for agent integration) - [POST https://wpnews.pro/mcp](https://wpnews.pro/mcp): MCP Streamable HTTP endpoint (2025-06-18 spec). JSON-RPC 2.0 over a single POST. Implements `initialize`, `tools/list`, `tools/call`, `resources/list`, `resources/read`, `prompts/list`, `prompts/get`. - [Server card](https://wpnews.pro/.well-known/mcp/server-card.json): SEP-2127 discovery doc with full tool/resource/prompt manifest. - Tools: `search`, `latest`, `topic`, `entity`, `article`, `list_topics`, `trending`, `stats`. - Resources: `wpnews://feed/latest`, `wpnews://corpus/llms-full`, `wpnews://corpus/index`. - Prompts: `summarize_today`, `find_emerging_trends`, `compare_entities`. ## Docs - [LLM Agent Integration Guide](https://wpnews.pro/for-llms): How to use wpnews.pro as a live news source for AI agents — covers MCP, OpenAI function calling, LangChain, RAG pipelines, and streaming formats. - [API Documentation](https://wpnews.pro/api): Full endpoint reference with interactive playground, code examples, and pricing. - [Methodology](https://wpnews.pro/methodology): How articles are sourced, processed, and translated. - [About](https://wpnews.pro/about): Editorial policies, data freshness, coverage scope. ## API - [GET /api/v1/context](https://api.wpnews.pro/api/v1/context?format=text&hours=24&limit=10): Pre-formatted news context block for direct system_prompt injection. Supports text, markdown, and json output formats. - [GET /api/v1/articles](https://api.wpnews.pro/api/v1/articles?lang=en&limit=10): Latest articles with title, summary, topics, entities, sentiment, and URL. - [GET /api/v1/search](https://api.wpnews.pro/api/v1/search?q=OpenAI&lang=en): Full-text search (FTS5 BM25) over all articles. - [GET /api/v1/digest/agent](https://api.wpnews.pro/api/v1/digest/agent?lang=en&hours=24): Compact digest with estimated_tokens, top_entities, top_topics for agents. - [GET /api/v1/articles/stream](https://api.wpnews.pro/api/v1/articles/stream?lang=en&limit=50): NDJSON stream — one article per line, ideal for vector DB ingestion. - [GET /api/v1/demo](https://api.wpnews.pro/api/v1/demo): Keyless demo — 3 live AI-summarized articles + trending entities. No API key required (50 calls/day/IP). Use to verify AI output quality before signing up. - [GET /api/v1/entities/trending](https://api.wpnews.pro/api/v1/entities/trending?lang=en&days=7): Companies and organizations surging in AI news coverage. - [GET /api/v1/openai-tools.json](https://api.wpnews.pro/api/v1/openai-tools.json): Ready-to-paste OpenAI function calling tools array (75 tools). Works with any OpenAI-compatible provider. - [GET /api/v1/anthropic-tools.json](https://api.wpnews.pro/api/v1/anthropic-tools.json): Same 75 tools in Anthropic tool_use format (input_schema, no type wrapper). Use with claude-opus-4-7 or claude-sonnet-4-6. - [GET /api/v1/adaptive-context](https://api.wpnews.pro/api/v1/adaptive-context): Auto-selects context depth (50-400 tokens) based on news activity. quiet=50 tokens, normal=150 tokens, surge=400 tokens. The API decides the budget — use context_text directly as system prompt. - [GET /api/v1/pulse](https://api.wpnews.pro/api/v1/pulse): Ultra-compact AI ecosystem pulse (~50-80 tokens): activity signal (surge/normal/quiet), activity_score, mood, top topic, top entity, best story. Use as a decision gate before loading /brief. - [GET /api/v1/brief](https://api.wpnews.pro/api/v1/brief?hours=24&articles=5): Unified smart briefing — top topics, trending entities, recent articles, and a compact text summary in one call. ~300-500 tokens. Ideal for agent system prompt injection. - [GET /api/v1/articles/featured](https://api.wpnews.pro/api/v1/articles/featured?hours=24&limit=5): Top articles ranked by entity+topic density (quality_score = entity_count*2 + topic_count). Ideal for curated agent picks and editorial highlights. - [GET /api/v1/articles/hot](https://api.wpnews.pro/api/v1/articles/hot?hours=6&limit=5): Breaking news radar — articles scored by hot_score = quality_score / log(1 + hours_ago). Surfaces the most important new stories before they trend. Poll every 30-60 min; compare slugs to detect new entries. - [GET /api/v1/insights/weekly](https://api.wpnews.pro/api/v1/insights/weekly?lang=en): Weekly intelligence brief — WoW article change, top topic/entity, sentiment shift, coverage gaps, highlight articles. - [GET /api/v1/articles/archive](https://api.wpnews.pro/api/v1/articles/archive?lang=en): Monthly archive index with article counts per month. - [GET /api/v1/articles/archive/{year}/{month}](https://api.wpnews.pro/api/v1/articles/archive/2026/01?limit=50): Paginated monthly article batch — up to 200 articles for training data or research. - [GET /api/v1/keywords/trending](https://api.wpnews.pro/api/v1/keywords/trending?lang=en&days=7): Trending keywords and bigrams from recent article titles. Useful for content intelligence agents. - [GET /api/v1/bulletin](https://api.wpnews.pro/api/v1/bulletin?lang=en&hours=24&items=5): Ultra-compact AI news bulletin as numbered list. ~100-200 tokens. Ideal for agent briefings and voice interfaces. - [GET /.well-known/agent.json](https://wpnews.pro/.well-known/agent.json): Google A2A Agent Card — machine-readable agent capability declaration for AI orchestrators. - [GET /article/{slug}.txt](https://wpnews.pro/article/sample.txt): Plain text article for LLM context injection. - [GET /article/{slug}.md](https://wpnews.pro/article/sample.md): Markdown with YAML frontmatter for document loaders. ## Research-friendly endpoints (NEW — for AI research agents) - [/research/topic/](https://wpnews.pro/research/topic/large-language-models): Topic research brief — 30-day timeline, top entities, sources. Drop-in LLM context. - [/research/entity/](https://wpnews.pro/research/entity/Anthropic): Entity dossier — all mentions, co-occurring entities, topic distribution. - [/research/timeline?topic=X&days=N](https://wpnews.pro/research/timeline?days=14): Chronological event timeline. - [/ai-events/today.md](https://wpnews.pro/ai-events/today.md): Last-24h snapshot of AI events. Refreshed every 30 minutes. ## Polling-friendly endpoints (NEW — for high-frequency agents) - [/feed/heartbeat](https://wpnews.pro/feed/heartbeat): ~50-byte response with last article id + timestamp. Recommended polling: 15s. - [/feed/latest.json](https://wpnews.pro/feed/latest.json): Latest 10 articles. Supports If-None-Match → 304. Recommended polling: 60s. - [/feed/since/](https://wpnews.pro/feed/since/0): Incremental feed — only articles newer than id. Recommended polling: 30s. ## Academic / harvester endpoints - [/oai?verb=Identify](https://wpnews.pro/oai?verb=Identify): OAI-PMH 2.0 — Semantic Scholar, BASE, OpenAIRE compatible. - [/oai?verb=ListRecords](https://wpnews.pro/oai?verb=ListRecords): Dublin Core metadata records. - [/stats/bots](https://wpnews.pro/stats/bots): Public AI bot activity dashboard. Proof of agent adoption. ## Optional - [MCP Server (HTTP+SSE)](https://wpnews.pro/for-llms#mcp): Native Model Context Protocol server. Connect via mcp-remote or use npx wpnews-mcp for stdio transport. - [JSON Feed](https://api.wpnews.pro/feed/json): JSON Feed 1.1 — machine-readable feed with full metadata. - [RSS Feed](https://api.wpnews.pro/feed/rss): RSS 2.0 — universal feed reader format. - [OpenAPI Schema](https://api.wpnews.pro/openapi.json): Complete OpenAPI 3.0 specification for code generation. - [Topics Index](https://wpnews.pro/topics): All topic categories with article counts. - [Entity Pages](https://wpnews.pro/entities): All tracked companies and organizations. - [/for-llms (full guide)](https://wpnews.pro/for-llms): Comprehensive integration guide with code examples.