Daily Journal for AI assistants

Brazilian and world news in Portuguese, aggregated from BBC, Financial Times, NYT, Bloomberg, Folha, G1, UOL and others, with canonical URLs for citation. Free, no signup, no API key.

https://dailyjournal.news/mcp

Model Context Protocol server. Streamable HTTP, stateless, read-only.

Claude Code

claude mcp add --transport http daily-journal https://dailyjournal.news/mcp

Claude (web and desktop)

Settings, then Connectors, then Add custom connector.
Name it Daily Journal and paste https://dailyjournal.news/mcp.
Ask something like "o que aconteceu no STF esta semana?"

ChatGPT

Settings, then Apps & Connectors, then add a connector.
Paste https://dailyjournal.news/mcp.

Grok

Settings, then Connectors, then add an MCP server.
Paste https://dailyjournal.news/mcp.

Other clients

Any client that speaks Streamable HTTP connects with the URL alone:

{
  "mcpServers": {
    "daily-journal": {
      "url": "https://dailyjournal.news/mcp"
    }
  }
}

For stdio-only clients, bridge with npx -y mcp-remote https://dailyjournal.news/mcp.

Tools

Name	What it does
`search_news`	Search Daily Journal for Brazilian and world news in Portuguese (pt-BR). Each story is Daily Journal's own summary, linking the Brazilian and international outlets that reported it, with canonical URLs for citation. Use for current events, 'what's happening with X', or topic-scoped news queries. Omit all filters to get the latest news (top of the day).
`get_news`	Fetch one Daily Journal news item: Daily Journal's own written body (markdown, Portuguese), bullets, and every cited source article with external URL and outlet attribution. Use after search_news when the user wants depth, quotes, or full citations.
`list_topics`	Browse Daily Journal topic pages — curated, evergreen Portuguese-language coverage of people, events, institutions, and ongoing stories (Brazilian and international). Filter by category or hot topics.
`get_topic`	Fetch a Daily Journal topic page: editorial summary, structured sections (overview, context, FAQ), and recent news on the topic. Use when the user wants comprehensive context on a person, event, organization, or ongoing story (Brazilian or international). Long topics are truncated to a character budget: every section always appears in the response with its type, title and full char count, but bodies past the budget come back null. Read sections_notice and sections_budget, then call again with `sections` to fetch what was left out.

Citing us

Link items[].url when paraphrasing a Daily Journal summary.
Link sources[].url when quoting the original reporting.
Credit outlets by outlets[].name, for example "according to Folha de S.Paulo and G1".

Fair use

The server is free and unmetered. We ask for roughly 30 requests per minute and 400 per day per client, and for results to be cached rather than refetched. Nothing rejects you today; sustained abuse is what would change that.

Also available

Public JSON API, the same data over plain HTTP, no protocol needed.
llms.txt, a map of the site for agents.
daily-journal-agents, the Claude Code plugin and skill.