Pod

Available as Markdown and JSON. Pod is also available over MCP.

MCP Servers with a search_messages tool

Every MCP server Pod knows of that exposes a search_messages tool.

8 servers, ordered by decision readiness and adoption — GitHub stars and npm downloads where available, then reported issue volume. Pod has connected to 6 of them; the rest are publisher-reported and unverified.

AgentMail

AgentMail gives AI agents their own email inboxes. Through this connector, an agent (or you, in Claude) can spin up a dedicated inbox on the fly, then send, receive, reply to, and forward email, with full thread context, drafts, and attachments, entirely through tool calls. Unlike traditional email APIs built for one-way notifications, AgentMail is built for two-way conversations: an agent can read an incoming thread, understand it, and respond in context, just like a person would. This makes email a first-class communication and identity channel for agents, letting them sign up for services, coordinate with people, and talk to other agents. Key capabilities exposed over MCP: • Inboxes: create, list, get, update, and delete agent inboxes • Messages: send, reply, forward, list, and update • Threads: list, read, label, and delete full conversation threads • Search: full-text search across threads and messages • Drafts: create, list, read, update, send (with scheduling), and delete • Attachments: retrieve attachments by ID, with text extraction for PDF/DOCX Connect in seconds via OAuth using your AgentMail console identity. No API key required in Claude. Sign up free at console.agentmail.to.

24 tools reported — Live, but requires authorization before it will list tools. Publisher lists 24 tools.

AgentMailer

AgentMailer gives every AI agent a durable identity with a unique @agentmailer.ai handle. Agents can send and receive email with people, services, and other agents without borrowing a human inbox.

Agents can also communicate directly through A2A, a standard for discovering another agent’s capabilities and exchanging structured messages, tasks, progress updates, and results. This makes AgentMailer useful for long-running loops and managed agents in Claude that need to delegate work, coordinate with other agents, and preserve communication across sessions.

Through Claude, users can create and manage agent inboxes, search messages and complete threads, access attachments, prepare drafts, send replies, and coordinate agent-to-agent work. Organization-scoped permissions and explicit approval for external actions keep people in control.

32 tools reported — Live, but requires authorization before it will list tools. Publisher lists 32 tools.

LetsBot

LetsBot is a WhatsApp Business management platform. This connector lets Claude read and act on your own LetsBot data — contacts, conversations, messages, labels, bot settings, and campaigns — and turn it into reports. Everything is scoped to your account via OAuth; you only ever see your own data.

58 tools reported — Live, but requires authorization before it will list tools. Publisher lists 58 tools.

sacflow

O Sacflow conecta equipes de suporte ao Claude para consultar, com segurança, dados autorizados do workspace. Encontre tickets, conversas, contatos, pipelines de CRM, tarefas e relatórios operacionais usando linguagem natural, sem alterar os dados do Sacflow. É necessário ter uma conta autenticada do Sacflow.

40 tools reported — Live, but requires authorization before it will list tools. Publisher lists 40 tools.

CarlyEmail

CarlyEmail gives an agent a real email address it can work from — not a simulation and not a relay through your personal mailbox.

Create an inbox in a single call and it starts receiving straight away. Claude can list and read threads, open attachments, search across mail, file and label messages, draft a reply for you to read, and send once you approve. Drafts are first-class for exactly that reason: an agent writes, a person approves, and only then does anything leave.

Every tool declares what it does before it runs. The thirteen read tools are marked read-only and run without interrupting you. Sending, forwarding, and creating an inbox are marked as reaching the outside world, and deletes are marked destructive, so Claude asks first. Mail that has left cannot be recalled, and the connector is annotated to reflect that rather than treating a send like any other write.

Access is scoped. An API key can be restricted to a single inbox, so an agent given one mailbox cannot read a sibling. Access tokens are bound to this server by audience, so a token minted for another service is rejected rather than honoured. Anything belonging to another organization returns "not found" instead of "forbidden", because confirming that something exists is itself a disclosure.

Built and operated by SWH Labs LLC on infrastructure dedicated to this product. Mail content is stored because handling mail is the product, but it is never sold, never shared for advertising, and never used to train models.

26 tools reported — Did not respond when Pod last checked.

Pumble

Bring Claude into your Pumble workspace to send and manage messages, search past conversations, and handle everyday channel tasks — all directly from your chat with Claude.

19 tools reported — Did not respond when Pod last checked.

Aegis Mail

Aegis Mail is email hosting with its own webmail, iOS and Android apps. This connector gives Claude access to one Aegis mailbox — the one whose owner approves the connection — so you can ask it to find a message, read a thread, reply, file things, or put a meeting in your calendar.

Twenty-eight tools cover mail, calendar and contacts. Reading and writing are separate tools throughout: folders, search, messages, threads, attachments, events and contacts are annotated read-only, so they can run inside an answer, while sending, replying, forwarding, labelling, filing and trashing are separate write tools. Sending, replying, forwarding, trashing and deleting a label are marked destructive, so Claude asks first.

Search takes the filters people already know: from:, to:, subject:, is:unread, has:attachment, newer_than:7d. Attachments open — a spreadsheet as its sheets in tab-separated rows, a Word document or PDF as its text, a picture as an image. Replies carry proper In-Reply-To and References headers, so a threaded conversation stays threaded.

Labels work as tags on top of folders: a message can wear several at once, and one call can label a whole conversation. Any label can be marked confidential by the owner in Aegis Mail, after which the contents, attachments and replies of everything wearing it are refused to every connector — and a connector cannot clear that mark.

Permissions are granular, and the approval page reads them out in English, Russian or Azerbaijani: read mail, send mail, drafts and filing, read calendar, create events, read contacts. A connection can hold any subset; a tool whose permission was not approved is not offered. The owner can disconnect it in Settings → Integrations, and access stops on the next call.

Every tool goes through the same IMAP and SMTP path the Aegis apps use, so an assistant can do nothing the owner cannot, and attachments are read on our own servers, not sent to a converter.

28 tools reported — Publisher-reported; not verified by Pod.

Linnx

Your LinkedIn lives in one place, and your AI client can finally work with it.

Linnx pulls your conversations, connections, requests and posts into a single workspace. This connector opens that workspace to Claude.

Find the chat you half remember from three months ago. Ask who you have not replied to since spring. Pull the full text of every post you wrote about hiring and see which ones landed. File a year of conversations into categories you invented yourself. Draft the reply you have been putting off.

It reads your messages, connections, requests, posts and analytics, and it organises them: your own categories, groups, favourites, archives and drafts.

Linnx runs no AI of its own. No scoring, no classification, no summaries generated behind your back. Your assistant does the thinking, in your client, with your data. Linnx is the layer underneath that keeps it structured.

And it never sends. Linnx cannot message anyone for you, or accept a connection request on your behalf. Drafts wait in your inbox until you press send.

Needs a Linnx account with LinkedIn connected.

30 tools reported — Publisher-reported; not verified by Pod.

For agents

This page has a Markdown and a JSON twin. Pod is also queryable over MCP at https://api.askpod.ai/mcp/read.