Skip to content

Self-improving AI Employees for everyone

docs: move Quick Start and install above the demo videos

trohitga58a50cdcb20
diff --git a/README.md b/README.md
index 6ff887ef..c27920e6 100644
--- a/README.md
+++ b/README.md
@@ -9,17 +9,31 @@ Your own AI assistant that does real work. Drag, drop, and connect AI agents to
No code required. No subscription. No usage limits. Bring your own API keys (or run models locally with Ollama / LM Studio for free).
-## How It Works
+## Quick Start
-[![How MachinaOS Works](docs/diagrams/how-it-works.svg)](https://raw.githubusercontent.com/trohitg/MachinaOS/main/docs/diagrams/how-it-works.svg)
+**Prerequisites:** Node.js 22+, Python 3.12+
-Pick nodes from the palette, drag them onto a canvas, connect them with lines, give your AI agent some memory and skills, and hit Play. Or **deploy** the workflow so it runs forever in the background — waiting for emails, responding to messages, checking in on a schedule, doing the work you'd rather not.
+```bash
+npm install -g machinaos
+machina start
+```
-## Three Workflows You Get on Day One
+Open http://localhost:3000 and click **Credentials** to connect your first AI provider.
-[![Default workflows that ship with MachinaOS](docs/diagrams/default-workflows.svg)](https://raw.githubusercontent.com/trohitg/MachinaOS/main/docs/diagrams/default-workflows.svg)
+<details>
+<summary><b>Run from source (for contributors)</b></summary>
-The first time you open MachinaOS, three example workflows load automatically. Open them on the canvas to see exactly how the pieces fit together, then edit any node and save your own version.
+```bash
+npm install -g pnpm
+git clone https://github.com/trohitg/MachinaOS.git
+cd MachinaOS
+pnpm install
+pnpm run dev
+```
+
+The `dev` task starts the Python backend, Vite client, WhatsApp service, and Temporal in parallel. See [SETUP.md](docs-internal/SETUP.md) and [SCRIPTS.md](docs-internal/SCRIPTS.md) for details, and [CONTRIBUTING.md](CONTRIBUTING.md) for the codebase map and contribution recipes.
+
+</details>
## See It In Action ↓
@@ -29,16 +43,17 @@ https://github.com/user-attachments/assets/5ee81bb3-12cf-4755-8532-7470c6f1d841
https://github.com/user-attachments/assets/5798fe61-8d26-4d3a-90aa-189bf4eec79f
-## Quick Start
+## How It Works
-**Prerequisites:** Node.js 22+, Python 3.12+
+[![How MachinaOS Works](docs/diagrams/how-it-works.svg)](https://raw.githubusercontent.com/trohitg/MachinaOS/main/docs/diagrams/how-it-works.svg)
-```bash
-npm install -g machinaos
-machina start
-```
+Pick nodes from the palette, drag them onto a canvas, connect them with lines, give your AI agent some memory and skills, and hit Play. Or **deploy** the workflow so it runs forever in the background — waiting for emails, responding to messages, checking in on a schedule, doing the work you'd rather not.
-Open http://localhost:3000 and click **Credentials** to connect your first AI provider.
+## Three Workflows You Get on Day One
+
+[![Default workflows that ship with MachinaOS](docs/diagrams/default-workflows.svg)](https://raw.githubusercontent.com/trohitg/MachinaOS/main/docs/diagrams/default-workflows.svg)
+
+The first time you open MachinaOS, three example workflows load automatically. Open them on the canvas to see exactly how the pieces fit together, then edit any node and save your own version.
## What You Can Build
@@ -155,23 +170,6 @@ Every LLM call and API request is tracked with USD cost. See per-provider spend
If anything goes wrong, the [Discord](https://discord.gg/c9pCJ7d8Ce) community is the fastest way to get help.
-## Run From Source
-
-<details>
-<summary><b>For contributors</b></summary>
-
-```bash
-npm install -g pnpm
-git clone https://github.com/trohitg/MachinaOS.git
-cd MachinaOS
-pnpm install
-pnpm run dev
-```
-
-The `dev` task starts the Python backend, Vite client, WhatsApp service, and Temporal in parallel. See [SETUP.md](docs-internal/SETUP.md) and [SCRIPTS.md](docs-internal/SCRIPTS.md) for details, and [CONTRIBUTING.md](CONTRIBUTING.md) for the codebase map and contribution recipes.
-
-</details>
-
## For Developers
Want to add a node, LLM provider, skill, or integration? One Python file = one node. The backend owns all the schemas; the frontend renders from them automatically. No frontend code required for most extensions.