Connect your assistant
This briefing publishes its archive as a Model Context Protocol server, so your own assistant can read it directly — every issue, the running story threads, the sources with their bias ratings, and the cost of each run.
Answers come from the archive itself, not from a search index built over it: ask for an issue by date and you get that issue. Read-only, public data, no key, no account.
Claude CodeRun in your terminal
claude mcp add --transport http news-digest https://news-digest.seanfloyd.dev/mcpCodexRun in your terminal
codex mcp add news-digest --url https://news-digest.seanfloyd.dev/mcpCursorOne-click install
Add to Cursor →VS CodeRun in your terminal
code --add-mcp "{\"name\":\"news-digest\",\"type\":\"http\",\"url\":\"https://news-digest.seanfloyd.dev/mcp\"}"Any other clientPaste this URL
https://news-digest.seanfloyd.dev/mcpNo client? Every tool is also a plain URL — start at the tool catalogue, or read the endpoint's own listing.
What your assistant can call
- get_issue
- One dated issue of the briefing as Markdown, by its YYYY-MM-DD date (from list_issues or search_headlines). Use get_latest_issue for the newest one.
- get_latest_issue
- The newest issue of the briefing as Markdown: every story with its headline, summary, why it matters, and the sources it was written from with their bias labels. Use this for 'what is in today's briefing' or 'what happened today'.
- get_sources
- Every feed the briefing reads, with its Media Bias/Fact Check bias and factuality rating, home region, and the perspective it was chosen to bring. Use this to interpret the bias labels on a story or to answer 'where does the briefing get its news'.
- get_stats
- Transparency statistics for a window of days, as JSON: per-source fetch health, how often each source was used at each tier, recent runs with articles kept and AI cost, and the de-duplication filter's numbers. Default window 30 days.
- get_thread
- One story thread's full history, newest installment first: the label, status, open questions the briefing is still watching, and for each day the matched headline and what was new. Ids come from list_threads.
- list_issues
- The archive index, newest first: each issue's date, its one-line preview, and a link to its Markdown. Use this to find which dates exist before calling get_issue.
- list_threads
- The briefing's running story threads: stories it has followed across several issues. Active threads first, then the most recently concluded, each with its id, label, status, and latest development. Use get_thread for a thread's full history.
- search_headlines
- Full-text search over every headline the briefing has published, up to 50 most relevant first, with the date and tier of each. Use this to find when a topic was covered, then get_issue for the full story.