Preliminary Notices via MCP

Construction-industry preliminary notices have to ship as physical mail with proof of delivery. Here's how to give an agent the ability to send them.

legalconstructionuse-case

· Updated

Why physical mail still matters here

In US construction, mechanic's lien rights typically depend on serving preliminary notices on the property owner, GC, and lender by certified mail (or equivalent). Email doesn't satisfy the statute. Physical mail with tracking does.

What the agent does

  • Pulls project, owner, GC, and lender data from your records.
  • Drafts the preliminary notice using the template that matches the project's state.
  • Calls send_letter for each recipient with the appropriate certified-mail flag.
  • Records the tracking IDs back into your project record for the audit trail.

Why MCP makes sense

A construction-PM-style agent already has tools for project data, contacts, and document templates. Adding Mailsnail gives it the missing physical-world action. The agent stops being a draft-and-remind tool and starts being end-to-end.

What we don't do

We don't write the legal content for you. Use the right state-specific template; we just print and mail it. If you need a richer compliance layer (deadline calculation, statutory language), keep that in your application and call us at the dispatch step.

Wire Mailsnail into your agent

Add one URL and sign in. Nothing to install and no API key. In Claude Code:

Claude Code
claude mcp add -t http mailsnail https://api.mailsnail.dev/mcp

Then run /mcp and choose Authenticate. In Claude Desktop or ChatGPT, add https://api.mailsnail.dev/mcp as a custom connector. Cursor and other JSON-config clients take:

mcp.json
{
  "mcpServers": {
    "mailsnail": {
      "url": "https://api.mailsnail.dev/mcp"
    }
  }
}

A one-page letter is from $1.21, and preview_letter shows the proof and exact price before anything is charged. Setup for every app →

See also

Automated Thank-You Notes with AI

An AI agent that decides who deserves a thank-you note, drafts it in your voice, and drops it in the mail. Setup, prompt patterns, and quality controls.

Agent Direct Mail

What direct mail looks like when an AI agent is the marketer: targeted, one-piece-at-a-time, decided per-event. Patterns, examples, and how to set it up.