Copy page

Social Media API SDKs in 8 Languages

Official client libraries for the OmniSocials API in 8 languages. Every SDK has the same shape: resource-based clients (client.posts.create(...)), typed requests and responses, automatic retries with backoff, a shared error hierarchy, and a webhook signature verification helper. Pick your language and you are posting in five lines.

Install

Each language has its own page with full examples: install, auth, posting, media, analytics, webhooks, and error handling.

Language	Package	Install	Docs
Node.js / TypeScript	@omnisocials/sdk	npm install @omnisocials/sdk	Node.js
Python	omnisocials	pip install omnisocials	Python
Go	github.com/OmniSocials/omnisocials-go	go get github.com/OmniSocials/omnisocials-go	Go
Ruby	omnisocials	gem install omnisocials	Ruby
Java	com.omnisocials:omnisocials-java	Maven / Gradle dependency	Java
PHP	omnisocials/omnisocials-php	composer require omnisocials/omnisocials-php	PHP
.NET	OmniSocials	dotnet add package OmniSocials	.NET
Rust	omnisocials	cargo add omnisocials	Rust

All 8 cover the full v1 surface: posts (including create-and-publish, per-platform options, and X / Bluesky / Mastodon threads), media (URL, base64, multipart, PDF carousels, presigned large-file uploads), folders, hashtag sets (CRUD plus applying a set at post create via hashtag_set / hashtag_placement), accounts, analytics (post, bulk, overview, accounts, best times), Instagram locations, Instagram Reel music search, and webhooks (CRUD plus secret rotation and signature verification).

Authentication

Every SDK reads the OMNISOCIALS_API_KEY environment variable, or takes the key in the constructor. Create a key in the OmniSocials app under Settings -> API. Keys start with omsk_live_ (production) or omsk_test_ (test mode).

Code
export OMNISOCIALS_API_KEY="omsk_live_<YOUR_KEY>"

Quickstart

Node.js / TypeScript

Code
import { OmniSocials } from "@omnisocials/sdk";

const client = new OmniSocials(); // reads OMNISOCIALS_API_KEY from env
const post = await client.posts.create({
  content: "Hello from the SDK",
  channels: ["instagram", "linkedin"],
  scheduled_at: "2026-08-01T09:00:00Z",
});

Python

Code
from omnisocials import OmniSocials

client = OmniSocials()  # reads OMNISOCIALS_API_KEY from the environment
post = client.posts.create(content="Hello from Python", channels=["instagram", "linkedin"])
print(post["data"]["id"])

A synchronous and an async client (AsyncOmniSocials) ship in the Python package. The Go, Ruby, PHP, Java, .NET, and Rust clients expose the same resources and methods adapted to each language's idiom. Full runnable examples for every resource live in each package's README (linked in the table above).

Configuration

Every client accepts the same options: apiKey, baseUrl, timeout (default 30s), and maxRetries (default 2). Retries use exponential backoff with jitter on 429, 5xx, and network errors, and honor the Retry-After header. Other 4xx responses are never retried.

Errors

Non-2xx responses raise a typed error from a shared hierarchy: a base error, an API error carrying status / code / body, and per-status subclasses (authentication, permission, not-found, validation, rate-limit with retry_after, server), plus a connection error and a webhook-verification error. See Rate limits and errors for the status codes.

Webhook verification

Each SDK ships a helper that verifies the X-OmniSocials-Signature header (HMAC-SHA256 over "{timestamp}.{rawBody}", constant-time compare, replay-window check) and returns the parsed event. Node example:

Code
import { verifyWebhookSignature } from "@omnisocials/sdk";

app.post("/webhooks/omnisocials", express.raw({ type: "application/json" }), (req, res) => {
  const event = verifyWebhookSignature({
    payload: req.body, // raw bytes, not parsed JSON
    signature: req.header("X-OmniSocials-Signature"),
    secret: process.env.OMNISOCIALS_WEBHOOK_SECRET,
  });
  // handle event.type ...
  res.sendStatus(200);
});

See Webhooks for event types and payloads.

Prefer a different integration?

• MCP Server for AI assistants (Claude, ChatGPT, Cursor, and more)
• Agent Skills for AI coding tools without MCP support
• API Reference for the raw HTTP endpoints