WaiverTrail + Claude (MCP): Send Waivers in Seconds — Setup Guide

If you run bookings, events, or guest lists, you already bounce between planning tools and your waiver system. Claude is where many teams draft messages and coordinate work. WaiverTrail is where templates, send limits, and signed records actually live. Connecting them means you can ask for status, list templates, and trigger sends in natural language instead of opening another tab for every small task.

This integration uses the Model Context Protocol (MCP), an open standard for tools over HTTPS, plus browser-based OAuth so you sign in to WaiverTrail the same way you would in the product. You do not need to paste long-lived secrets into Claude for the standard flow. For the full product overview of the connector, read the dedicated WaiverTrail + Claude integration page.

MCP is a structured way for Claude to call approved server capabilities, such as listing waiver templates or checking whether a guest has completed a form. Your organization stays in control through scopes you authorize during sign-in. Each request is authenticated to your WaiverTrail account rather than relying on a model-supplied tenant id.

Exact menu labels can change as Claude updates its UI, but the workflow is stable. In Claude, open Settings, find Integrations or custom connectors, and choose Add MCP server. Paste your production MCP URL. Many teams use the dedicated Claude path documented on the integration page, for example `https://waivertrail.com/api/mcp/claude` (replace the host if you use a custom domain).

Click Connect. A WaiverTrail browser tab should open for sign-in and authorization. Complete login, approve access, then return to Claude. Run a simple first prompt such as list my waiver templates to confirm tools are available. If you manage connectors inside WaiverTrail as well, you can open Settings with the integrations tab to review access from your account side.

Short, specific prompts work best. Try listing and choosing a template first, then asking for a send when your account allows the send scope. Examples you can adapt: List my active waiver templates and give a one-line summary of each. Has the guest with email example@example.com completed the waiver for Saturday check-in? Send the standard liability template to example@example.com with a short note that it is for booking ABC123.

Always confirm your plan, quotas, and scopes before relying on automated sends. The same business rules that apply inside WaiverTrail still apply when the request originates from Claude. For how pre-arrival delivery fits into operations, see booking and QR check-in. For template structure, see waiver templates.

Treat the connector like any other integration with least privilege. Remove the connector in Claude when someone leaves the team or you no longer need AI access. Avoid pasting sensitive personal data into chat unless your policies explicitly allow it; use booking identifiers or operational emails your workflow already treats as normal business data.

WaiverTrail is built for HTTPS, scoped access, and auditability. Claude only receives what your tool definitions and scopes expose. For a broader platform picture, read digital waiver software.

If authorization never appears, remove the connector in Claude and add it again with a fresh URL. Confirm pop-up blockers are off for the WaiverTrail domain. If Claude shows connected but tools fail, complete disconnect and reconnect after a full OAuth round trip. High-volume teams should also confirm firewalls or WAF rules allow Anthropic outbound traffic to your public HTTPS endpoint.

Ready to try the flow end to end? Start your free WaiverTrail trial, connect Claude using MCP, and run your first status or send prompt in minutes. Keep the Claude integration guide bookmarked for URL and step references as Anthropic updates the client UI.

Go deeper with topic-specific articles: OAuth, scopes, and data boundaries, event waiver workflow with MCP, MCP vs API keys for automation, and connector troubleshooting.