no ollama serve command

docs/docs.json

@@ -129,6 +129,7 @@
                   "/integrations/opencode",
                   "/integrations/droid",
                   "/integrations/goose",
+                  "/integrations/omp",
                   "/integrations/pi",
                   "/integrations/pool"
                 ]

docs/integrations/index.mdx

@@ -15,6 +15,7 @@ Coding assistants that can read, modify, and execute code in your projects.
 - [OpenCode](/integrations/opencode)
 - [Droid](/integrations/droid)
 - [Goose](/integrations/goose)
+- [OMP](/integrations/omp)
 - [Pi](/integrations/pi)
 - [Pool](/integrations/pool)
 

docs/integrations/omp.mdx

@@ -0,0 +1,47 @@
+---
+title: OMP
+---
+
+OMP is an AI coding agent with IDE integration that runs in your terminal.
+
+## Install
+
+Install [OMP](https://omp.sh):
+
+```bash
+curl -fsSL https://omp.sh/install | sh
+```
+
+On Windows, install from PowerShell:
+
+```powershell
+irm https://omp.sh/install.ps1 | iex
+```
+
+## Usage with Ollama
+
+OMP discovers Ollama automatically when the Ollama server is running locally.
+
+Start Ollama if it is not already running, then pull a model:
+
+```bash
+ollama pull qwen3.5
+```
+
+Then launch OMP:
+
+```bash
+omp
+```
+
+Use `/model` in OMP and select the Ollama model, such as `ollama/qwen3.5` and set it as the default.
+
+To launch OMP with a specific Ollama model:
+
+```bash
+omp --model ollama/qwen3.5
+```
+
+<Note>For coding agents, larger context windows work best. See [Context length](/context-length) for more information.</Note>
+
+For more OMP model configuration options, see the [custom models documentation](https://omp.sh/docs/custom-models).