Connect Everything: 80+ Actions, One Platform

Andy Esser
Apr 13, 2026

Connect Everything: 80+ Actions, One Platform

Part 3 of our Workflows Unleashed series. This week: why Flomation's actions aren't just HTTP wrappers with a logo.

The promise of every automation platform is the same: "we connect to everything." The reality is usually a list of logos on a marketing page, each backed by a generic HTTP connector wearing the service's colours.

Flomation takes a different approach. Each action is purpose-built for the API it talks to. It understands the data types, the authentication patterns, the error codes, and the response structures. The result is a node you can configure in seconds, not minutes — because it already knows what the API expects.

The Action Categories

Flomation ships with over 80 actions across 17 categories. Here's what that actually means.

Messaging

Slack speaks the Bot API natively. Channel IDs, thread replies, message formatting with mrkdwn (not Markdown — Slack's own dialect where *bold* works and **bold** renders literally). The action handles authentication, posting, and error responses.

Telegram uses the Bot API with support for message sending, reply threading, and Markdown formatting. Discord connects via webhooks for outbound messages. SMTP email handles HTML bodies, recipients, and delivery.

Each messaging action understands its platform's formatting, threading model, and character limits. You don't have to remember that Slack uses *bold* while Telegram uses **bold** — the action handles it.

Cloud

AWS S3 has four dedicated actions: list (with prefix filtering), get (download an object), put (upload), and delete. Each uses the standard AWS SDK credential chain — IAM roles in production, environment variables or credential files in development.

AWS EC2 describes instances with full metadata — state, type, IP addresses, tags. DynamoDB supports Query, Scan, PutItem, GetItem, and DeleteItem with full expression syntax.

DevOps

Git actions cover the full lifecycle: clone, branch, checkout, add, commit, push, pull, status, and tag. SSH key authentication is supported via configurable key paths. The Git polling trigger watches for branch changes without requiring CI webhook configuration.

HTTP Request is the universal adapter for anything without a dedicated action. Full header control, all methods (GET, POST, PUT, PATCH, DELETE), authentication (Bearer, Basic, custom headers), and structured response parsing.

Bash Script executes shell commands in a sandboxed working directory. When no built-in action exists for your tool, this is the escape hatch — and it's a proper one, with stdout capture, exit code handling, and environment variable injection.

Data Processing

Data Combine merges outputs from multiple parent nodes into a single object. Data Extract selectively picks fields from upstream data. Data Rename remaps keys between systems that use different naming conventions. Set Variable stores values accessible anywhere in the flow via ${var.name}.

These utility actions are the glue between integrations. They're small, fast, and composable — exactly what you need when System A returns user_email but System B expects email_address.

Why Purpose-Built Matters

A generic HTTP action can technically call any API. But it makes you do all the work: construct the URL, set the headers, format the body, parse the response, handle errors. For every single call. In every single flow.

A purpose-built action has already done that work. The Slack action knows about channel IDs and mrkdwn. The PostgreSQL action knows about connection strings and parameterised queries. The S3 action knows about bucket policies and content types.

The difference isn't capability — it's speed. You go from "I need to automate this" to "it's automated" in minutes, not hours.

When You Need Something We Don't Have

The HTTP Request action handles any REST API. The Bash Script action handles anything your server can run. Between the two, there's no API or tool that's out of reach.

And if you find yourself building the same HTTP request pattern repeatedly, that's a signal — it might be time for a new dedicated action. Flomation's action system is extensible by design.

Next Week

We go deep on databases — PostgreSQL, MySQL, Oracle, MongoDB, Redis, DynamoDB — and why each gets its own dedicated action instead of a shared SQL executor.

www.flomation.co — free to start, no credit card.