newsapi-ai-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| MCP_AUTH_TOKEN | No | HTTP mode only: bearer token clients must send | |
| NEWSAPI_AI_KEY | Yes | Your newsapi.ai API key (required) | |
| NEWSAPI_AI_HOST | No | API host, default `https://eventregistry.org` | |
| NEWSAPI_AI_MAX_DAYS | No | `7` or `31`: hard cap, never query older data (archive queries cost more tokens) |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| newsapi_search_articlesA | Search ~150k news/blog/press-release sources worldwide (newsapi.ai / Event Registry). Filter by keywords, concepts, categories, sources, locations, language and date. For entities (companies, people, places) prefer concept_uris over keywords – resolve them first with newsapi_suggest. Each call consumes API tokens; queries older than 31 days cost more. Returns compact results unless response_format=raw. |
| newsapi_get_articlesA | Fetch full details (complete body text, authors, concepts, categories) for articles by their newsapi.ai URI (from search results) or by original article URL. |
| newsapi_search_eventsA | Search events: clusters of articles (often across languages) reporting on the same real-world happening. Good for 'what happened around X' and for spotting stories with broad coverage. Same filters as article search. |
| newsapi_get_eventB | Details for one event (summary, concepts, location) plus optionally its top articles. |
| newsapi_suggestA | Autocomplete names to the URIs that the search tools need: concepts (people, organisations, places, topics → Wikipedia URI), sources (news outlets → domain URI), source_groups, categories, locations, authors. Call this BEFORE filtering by entity, source or category. |
| newsapi_usageA | Show how many newsapi.ai API tokens are available and used on the current plan. |
| newsapi_raw_requestA | Escape hatch: POST arbitrary parameters to any newsapi.ai endpoint under /api/v1/ (e.g. /api/v1/article/getArticles with resultType=conceptTrends, /api/v1/minuteStreamArticles, /api/v1/trends). The API key is added automatically. Prefer the dedicated tools when they fit. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 7 tools
Each tool has a distinct purpose: search_articles vs search_events target articles vs clusters, get_articles vs get_event fetch details for each, and suggest/usage/raw_request are clearly separate meta/helper tools. Boundaries are well-drawn, and raw_request is explicitly labeled as a fallback for cases the dedicated tools don't cover.
All tools share a clean newsapi_ prefix with snake_case and mostly verb_noun structure (search_articles, search_events, get_articles, get_event, suggest, raw_request). newsapi_usage is noun-only and breaks the verb pattern slightly, but overall the convention is predictable.
Seven tools is a well-scoped surface for a news search API, covering search, retrieval, discovery/autocomplete, usage, and an escape hatch without redundancy. Each tool earns its place.
The lifecycle is well covered: query (search_articles/events), retrieve full details (get_articles/get_event), resolve entities/sources (suggest), monitor quota (usage), and extend via raw_request. Some higher-level analytics (e.g. trends, concept trends) require the raw escape hatch rather than dedicated tools, a minor gap.