Back to browse
Hail.so – Send and receive emails, calls, SMS via a single MCP/API/CLI

Hail.so – Send and receive emails, calls, SMS via a single MCP/API/CLI

by r13i·Jun 29, 2026·1 point·0 comments

AI Analysis

●●●BangerBig BrainSolve My Problem

One MCP endpoint gives agents real phone numbers and email—no Twilio hell.

Strengths
  • Solves the agent communication gap with production-tested voicebot infrastructure
  • Unified event stream across calls, SMS, and email with native interruption handling
  • One-click MCP integration for Claude, ChatGPT, Cursor, and eight other clients
Weaknesses
  • Managed cloud option creates vendor lock-in concerns for self-hosters
  • Complex stack (Twilio, SES, LiveKit, ElevenLabs) means more failure points
Category
Target Audience

AI agent developers, automation engineers

Similar To

Twilio · Retell AI · Vapi

Post Description

Hello HN,

Hail is an MCP server, CLI and API for email, SMS and phone calls. Inbound and outbound.

It packages Twilio, AWS SES (with or without custom domains), LiveKit, ElevenLabs, Cartesia, and LLM providers under a consistent API.

The implementation comes from about a year of production work building industrial AI voicebots, and tooling for all sorts of AI agents.

Check the README for the current stack and milestones for what's coming.

Open source, self-hostable backend: https://github.com/hail-hq/hail/

How this project came about? ----------------------------

I spent over a year as CTO of an AI startup building a useful chatbot + voicebot for industrial companies, and the most useful features turned out to be:

> generating documents and sending them via emails > calling staff members to collect specific information like "where did you leave the key to the building" or "what's the specific values to set for the oil refining machine" > sending SMS reminders for staff to do XYZ

However it was pretty hard to get all 3 working well, and at a low cost: sending and receiving emails, phone calls, and sms.

It requires several days at best to connect Twilio SIP trunks to a server to handle them, then set up AWS SES to send emails, etc. and that is if you know what you're doing, and not starting from scratch.

Plus, each provider would give you a separate invoice you'd have to do bookkeeping for.

So I've put all the learnings to work in a single project, not only so that I don't loose the learning, but also because every company needs to give their AI agent or backend a way to communicate with the world via email, phone, or SMS.

A single API, MCP, and CLI.

Why not just set up my own Twilio + Resend + Postmark? ------------------------------------------------------

> separate integrations for email, voice and SMS. Takes very long to set up > deploying telephony requires SIP infrastructure > several dashboards, messy UI/UX > several bills to keep track off in bookkeeping > no native MCP server or CLI for AI agents

This project is not replacing existing providers (twilio, amazon ses, etc), but packages them into a single platform with sane defaults.

Where are we at? ----------------

Right now Hail has good support for outbound phone calls and emails, and the inbound emails is well underway for a final release using auto-forwarding and webhooks.

What is left is support for inbound phone calls, and SMS inbound/outbound.

What next? ----------

You can host the project yourself or check out https://hail.so/

Similar Projects

AI/ML●●Solid

Email for AI Agents

Solves the agent email gap with a simple API, competing directly with Resend and SendGrid.

Ship ItSolve My Problem
johnjoubert
111mo ago
Developer Tools●●Solid

AgenticMail – Email, SMS, and multi-agent coordination for AI agents

Hands-off: each agent gets its own mailbox, phone number, inbox and API key, and the project bundles a local Stalwart mail server in Docker plus Gmail-relay or custom-domain outbound with DKIM/SPF/DMARC. Concrete safety features (outbound scanning to catch API keys/PII) and Google Voice SMS extraction for verification codes are genuinely useful and uncommon together. If you want to prototype agents that actually talk to people/services this is a focused, pragmatic stack — but expect telephony fiddliness and compliance trade-offs in real deployments.

WizardryNiche Gem
agenticmail
214mo ago