mcpField
P
postbag

postbag

Local
postbag.devDeveloper Tools0Updated 5w ago
RepositoryVisit website

About postbag

Postbag is a form backend for people operating multiple sites and the AI agents that build them. It centralizes submissions and routes them through retryable delivery to email, Telegram, webhooks, partners, or CRMs.

Key features

  1. 01Collect and centralize website form submissions
  2. 02Map varied form fields to a shared shape
  3. 03Route data to email, Telegram, or webhooks
  4. 04Deliver through a durable outbox with retries

Use cases

  1. 01Combine forms from multiple sites into one stream
  2. 02Send normalized form data to a partner
  3. 03Retry delivery after downstream failures

MCP support

The README mentions an MIT-licensed `@postbag/mcp` client package but does not describe its MCP tools or connection method.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
  "mcpServers": {
    "dev-postbag": {
      "command": "npx",
      "args": [
        "-y",
        "@postbag/mcp"
      ],
      "env": {
        "POSTBAG_API_KEY": "<POSTBAG_API_KEY>",
        "POSTBAG_API_URL": "<POSTBAG_API_URL>"
      }
    }
  }
}
Run commandnpx -y @postbag/mcp
Environment variables to set
POSTBAG_API_KEYSecretA Postbag API key (pb_live_… or pb_test_…). Mint one at https://postbag.dev/app or with `postbag login`.
POSTBAG_API_URLBase URL of the Postbag API. Defaults to https://postbag.dev — set this only for a self-hosted instance.

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npm@postbag/mcp0.1.1

Tools

64

Saved tool list · updated 2w ago

This tool takes no parameters.

Details

From the project README

Postbag

CI npm: postbag npm: @postbag/sdk npm: @postbag/mcp License: AGPL-3.0 skills.sh

A form backend that routes. Point any HTML form at a Postbag endpoint and it lands in your inbox — then goes wherever it should: email, Telegram, a webhook, a partner, your CRM. Built for people who run many sites, and for the AI agents that build them.

<form action="https://api.postbag.dev/s/fm_8f3kq2" method="POST">
  <input name="email" type="email" required>
  <textarea name="message"></textarea>
  <button>Send</button>
</form>

That's the whole integration for one form. Everything else — grouping forty forms into one stream, mapping their different fields onto one shape, sending that shape to a partner between two dates, a daily digest, never losing a submission — is there when you need it and invisible when you don't.

New here (human or AI agent)? Read CLAUDE.md first.

Why it exists

Every website has forms. Every form needs somewhere to go. Formspree and friends solve "somewhere" for one site; they fall over when you have fifteen sites feeding one partner with fifteen slightly different forms. Postbag is the source of truth for what each form collects, what each downstream system receives, and the versioned contract between them.

The one idea

The database makes it correct; events make it fast.

A submission is accepted the instant it is a row. Delivery is a durable outbox that a worker drains with retries. Realtime nudges are an accelerator, never a transport. If every event stream in the system died, every submission would still arrive — late, never lost. (Inherited from the lead pipeline this generalises; see docs/ARCHITECTURE.md.)

Status

Phase 1 — MVP live at postbag.dev (site, docs, dashboard at /app, API at /v1). See PROGRESS.md for the live blueprint and docs/ROADMAP.md for phases.

Documentation

Start at docs/README.md.

License

The server, dashboard, site and domain packages (apps/*, packages/core, packages/db, packages/auth) are licensed under AGPL-3.0. The client packages that run inside your own code — @postbag/sdk, the postbag CLI and @postbag/mcp — are MIT. Rationale in ADR-006.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.