1comms

Agentic messaging substrate

The API built for engineers and autonomous agents.

Build custom communication clients, internal support terminals, and autonomous agent workflows on a unified schema covering seventeen platforms.

Philosophy & architecture

Our documentation, type definitions, and prompt files are designed for direct consumption by large language models. We publish llms.txt files and structured prompt packs alongside typed clients in TypeScript, Python, Swift, and Kotlin. Your coding agent can read our documentation, scaffold a complete custom interface, and run its first live test in under sixty minutes.

The unified API collapses seventeen closed messaging networks into one normalized schema. You listen to live events through durable websocket streams, receive verified webhooks, and query petabytes of historical context with a single API signature. You build your product logic once instead of wrestling with seventeen fragile third-party platform libraries.

Production routing runs directly through our core carrier-grade infrastructure. We provide strict delivery SLAs, idempotent outbound dispatch, and exactly-once delivery semantics across every bridge. Your application gains instant enterprise scale, automated token refreshes, and deep cold-storage search without hosting bespoke protocol connectors.

agent-terminal.ts

$ bun add @1comms/sdk

const comms = new OneComms(apiKey);

for await (const msg of comms.messages.stream()) {

const to = msg.sender.id;

await comms.messages.send({ to, network: "whatsapp", text: "Received." });

}

API surface

Six signatures. Seventeen networks.

Every endpoint speaks the same normalized schema, whatever sits on the other side of the bridge.

messages.stream

Subscribe to a live, multi-protocol websocket stream of inbound messages, read receipts, and typing states.

messages.send

Dispatch outbound messages, media attachments, and structured replies to any network through a single JSON payload.

search.query

Run natural language RAG queries and exact vector searches across years of conversation history in under fifty milliseconds.

contacts.merge

Programmatically unify disjointed phone numbers, social handles, and email addresses into single identity nodes.

media.transcribe

Extract text from inbound voice notes and parse visual data from image attachments via unified worker endpoints.

export.create

Generate structured, uncompressed JSON archives of message histories and raw attachments for offline processing.

For your agent

Paste our docs URL into Cursor or Claude Code and watch your agent build a messaging terminal in minutes.

https://1comms.b0rz.com/llms.txt