Ask your AI tool which trials sit next to your asset.

What you can ask

• "Which KIF18A trials are running in ovarian cancer, and who sponsors them?" The compare_trials tool returns up to three interventional studies side by side: sponsor, phase, status, declared arms, primary outcomes and last registry update.
• "What does the KRAS G12C trial landscape look like right now?" search_registry gives total and active study counts, leading sponsors, the phase mix and the most recently updated studies, with related Prognyx briefings.
• "Has Prognyx written about multiple myeloma CAR-T?" search_briefings searches the published briefings, newest first.

Use a target rather than a single drug name to see other programmes: "B7-H3 in small cell lung cancer" returns Phase 3 studies from GSK, Daiichi Sankyo and Qilu.

Connect it

Claude (desktop app and claude.ai)

Settings → Connectors → Add custom connector. Name it Prognyx and paste the server URL above.

Claude Code

claude mcp add --transport http prognyx https://www.prognyx.com/api/mcp

Cursor, Windsurf and other clients

Add the server to the client's MCP configuration file:

{
  "mcpServers": {
    "prognyx": { "url": "https://www.prognyx.com/api/mcp" }
  }
}

What it does not do

It reads the public registry and says what the records declare. It does not decide who competes with whom: a shared keyword is not a shared population, and a registry record is not an efficacy result. Counts for sponsors and phases come from a recent sample, stated in each answer. The registry is queried live, so results change as records are updated.

The reading that turns those records into a decision (which programme matters for your asset, and what would change your next move) is what a Prognyx pilot delivers.