@pipeworx/press-intel-pack

Connect: https://gateway.pipeworx.io/press-intel/mcp

Tools: 8

Tools

• match_story_to_journalists — Find the journalists most likely to care about a specific story, ranked, WITH the articles that prove each match. Give it the story itself — a paragraph of what happened and why it matters, not a keyw
• search_journalists — Find journalists in the corpus by name, outlet, or beat vertical. Returns their outlet, article count, beat profile and published contacts. Use match_story_to_journalists when you have a story — this
• get_journalist — Everything the corpus holds on one journalist: outlet and employment history, beat profiles across verticals, published contact points with provenance, recent articles, and any recorded outreach. Take
• search_articles — Search the press corpus by headline text, outlet, vertical or date. Returns headline, URL, outlet, publication date, our summary, and the bylined journalists. Use this to check what has already been w
• journalist_recent_coverage — What one journalist has published lately, newest first, with our summaries and the topics and entities we extracted. The question this answers before an outreach: are they still on this beat, and what
• create_media_list — Save a set of journalists as a named media list, keeping each one’s score and supporting evidence AS OF NOW. A list is a decision record — the beat profiles move nightly, so re-deriving the scores lat
• get_media_list — Read back a saved media list: its members with the score and evidence recorded when it was built, plus each journalist’s current contact points and beat. Pass no slug to list every saved list.
• record_outreach — Log that an outreach happened. Append-only and deliberately dumb: this pack does not send anything, it records what a human or another system already did, so the response history exists to learn from