Skip to content

Install the Free Edition

This page mirrors the package README.md in the public repo. If they diverge, the README is the source of truth.

Run it straight from npm — no clone required:

Terminal window
npx -y elicitly

Register it with any MCP host as a stdio server. In Claude Code:

Terminal window
claude mcp add elicitly -- npx -y elicitly

Or add it to a host config file (.mcp.json or client settings):

{
"mcpServers": {
"elicitly": { "command": "npx", "args": ["-y", "elicitly"] }
}
}

Download elicitly.mcpb from the latest release and open it in Claude Desktop (or Settings → Extensions → Install extension). The bundle is self-contained — no Node.js or npm setup of your own; it runs on Claude Desktop’s managed runtime. Build it from source with pnpm --filter elicitly build:mcpb.

Claude Desktop and dialogs: current Claude Desktop builds don’t advertise elicitation to local MCP servers (config-file and extension installs alike — measured, not assumed), so elicit_confirm and elicit_form return their graceful no-answer contract there rather than raising a dialog. elicit_doctor still reports the host truthfully — that report is the point. See the Elicitation Support Matrix for hosts where dialogs work, and the Pro Edition for approvals that reach a browser-based review page when the host has no native dialog.

The Free Edition is open source at github.com/elicitly/elicitly. To run it from a checkout, build the CLI and point your host at the compiled entry:

Terminal window
git clone https://github.com/elicitly/elicitly.git
cd elicitly
pnpm install
pnpm --filter elicitly build # produces packages/elicitly/dist/cli.mjs

Register the built file with an absolute path in your stdio host:

Terminal window
claude mcp add elicitly -- node /absolute/path/to/packages/elicitly/dist/cli.mjs

After editing the source, rebuild (pnpm --filter elicitly build) and restart the server in your host.

Try it now — paste this prompt to your agent:

Call the elicit_confirm tool with the question "Ready to try Elicitly?"

A native OK/Cancel dialog should appear in your host; on OK the tool returns { "confirmed": true }.

From there, ask for elicit_doctor — a passive capability report with no user prompt — then elicit_doctor with probeElicitation: true to run a real elicitation round-trip, and elicit_form to fire your own schema.

When the dialog isn’t enough — a decision that must survive the session, reach another device, or leave an audit trail — the same three calls plus the approval tools run hosted in the Pro Edition (pricing).