UI integrations

integrations/chainlit.md

@@ -0,0 +1,110 @@
+---
+layout: integration
+name: Chainlit
+description: Use Chainlit UI for your Haystack apps through Hayhooks
+authors:
+  - name: deepset
+    socials:
+      github: deepset-ai
+      twitter: deepset_ai
+      linkedin: https://www.linkedin.com/company/deepset-ai/
+pypi: https://pypi.org/project/hayhooks/
+repo: https://github.com/deepset-ai/hayhooks
+type: UI
+report_issue: https://github.com/deepset-ai/hayhooks/issues
+logo: /logos/chainlit.png
+version: Haystack 2.0
+toc: true
+---
+
+### Table of Contents
+
+- [Overview](#overview)
+- [Installation](#installation)
+- [Usage](#usage)
+- [License](#license)
+
+## Overview
+
+[Chainlit](https://chainlit.io/) is an open-source Python package for building 
+production-ready Conversational AI. By exposing your Haystack app (standalone agent or 
+pipeline) through [Hayhooks](https://github.com/deepset-ai/hayhooks) as 
+OpenAI-compatible endpoints, you can run the Chainlit chat UI inside your Hayhooks 
+server, giving you a zero-configuration frontend to interact with your deployed 
+pipelines without a separate client. 
+
+For full details, see the [Hayhooks Chainlit integration guide](https://deepset-ai.github.io/hayhooks/features/chainlit-integration).
+
+## Installation
+
+Install Hayhooks with the `chainlit` extra:
+
+```bash
+pip install "hayhooks[chainlit]"
+```
+
+## Usage
+
+### Hayhooks Quick Start
+
+The simplest way to enable the Chainlit UI is via the `--with-chainlit` flag:
+
+```bash
+hayhooks run --with-chainlit
+```
+
+This starts Hayhooks with the embedded Chainlit UI available at `http://localhost:1416/
+chat`.
+
+### Create a Pipeline Wrapper
+
+```python
+# pipelines/my_chat/pipeline_wrapper.py
+from typing import Generator
+
+from haystack import Pipeline
+from haystack.components.builders import ChatPromptBuilder
+from haystack.components.generators.chat import OpenAIChatGenerator
+from haystack.dataclasses import ChatMessage
+
+from hayhooks import BasePipelineWrapper, streaming_generator
+
+class PipelineWrapper(BasePipelineWrapper):
+    def setup(self) -> None:
+        self.system_message = ChatMessage.from_system("You are a helpful assistant.")
+        chat_prompt_builder = ChatPromptBuilder()
+        llm = OpenAIChatGenerator(model="gpt-4o-mini")
+
+        self.pipeline = Pipeline()
+        self.pipeline.add_component("chat_prompt_builder", chat_prompt_builder)
+        self.pipeline.add_component("llm", llm)
+        self.pipeline.connect("chat_prompt_builder.prompt", "llm.messages")
+
+    def run_chat_completion(self, model: str, messages: list[dict], body: dict) -> Generator:
+        chat_messages = [self.system_message] + [
+            ChatMessage.from_openai_dict_format(msg) for msg in messages
+        ]
+        return streaming_generator(
+            pipeline=self.pipeline,
+            pipeline_run_args={"chat_prompt_builder": {"template": chat_messages}},
+        )
+```
+
+Pipelines must support chat completion (e.g. using `streaming_generator` or `async_streaming_generator`). See [OpenAI compatibility](https://deepset-ai.github.io/hayhooks/features/openai-compatibility) and the [pipeline examples](https://deepset-ai.github.io/hayhooks/examples/overview/) for implementation details.
+
+### Run Hayhooks with UI
+
+```bash
+hayhooks run --with-chainlit --pipelines-dir ./pipelines
+```
+
+Navigate to [`http://localhost:1416/chat`](http://localhost:1416/chat) in your browser. You'll see your deployed pipeline and can start chatting!
+
+![Chainlit UI](../images/hayhooks-chainlit.gif)
+
+### Examples
+Learn how to build an end-to-end agent example with Haystack, Hayhooks and Chainlit in [Chainlit Weather Agent Example](https://github.com/deepset-ai/hayhooks/tree/main/examples/pipeline_wrappers/chainlit_weather_agent)
+
+## License
+
+`hayhooks` and `chainlit` are distributed under the terms of the [Apache-2.0](https://spdx.org/licenses/Apache-2.0.html) license.

integrations/openwebui.md

@@ -0,0 +1,73 @@
+---
+layout: integration
+name: Open WebUI
+description: Use Open WebUI as a chat frontend for your Haystack apps through Hayhooks
+authors:
+  - name: deepset
+    socials:
+      github: deepset-ai
+      twitter: deepset_ai
+      linkedin: https://www.linkedin.com/company/deepset-ai/
+pypi: https://pypi.org/project/hayhooks/
+repo: https://github.com/deepset-ai/hayhooks
+type: UI
+report_issue: https://github.com/deepset-ai/hayhooks/issues
+logo: /logos/openwebui.png
+version: Haystack 2.0
+toc: true
+---
+
+### Table of Contents
+
+- [Overview](#overview)
+- [Installation](#installation)
+- [Usage](#usage)
+- [License](#license)
+
+## Overview
+
+[Open WebUI](https://openwebui.com/) is an open-source chat UI for LLM apps. By exposing your Haystack pipelines and agents through [Hayhooks](https://github.com/deepset-ai/hayhooks) as OpenAI-compatible endpoints, you can use Open WebUI as the frontend: run Hayhooks and Open WebUI (separately or via Docker Compose), then connect Open WebUI to Hayhooks in Settings. You get streaming, optional [status and notification events](https://deepset-ai.github.io/hayhooks/features/openwebui-integration#open-webui-events), and optional [OpenAPI tool server](https://deepset-ai.github.io/hayhooks/features/openwebui-integration#openapi-tool-server) integration.
+
+For full details, see the [Hayhooks Open WebUI integration guide](https://deepset-ai.github.io/hayhooks/features/openwebui-integration).
+
+## Installation
+
+Install Hayhooks:
+
+```bash
+pip install hayhooks
+```
+
+Install and run Open WebUI separately, e.g. with Docker:
+
+```bash
+docker run -d -p 3000:8080 --add-host=host.docker.internal:host-gateway -e WEBUI_AUTH=False -v open-webui:/app/backend/data --name open-webui ghcr.io/open-webui/open-webui:main
+```
+
+For a pre-wired setup, use the [Hayhooks + Open WebUI Docker Compose](https://github.com/deepset-ai/hayhooks-open-webui-docker-compose) (see [Quick Start with Docker Compose](https://deepset-ai.github.io/hayhooks/getting-started/quick-start-docker/)).
+
+## Usage
+
+### Connect Open WebUI to Hayhooks
+
+1. Start Hayhooks with your pipelines:
+
+   ```bash
+   hayhooks run --pipelines-dir ./pipelines
+   ```
+
+2. In Open WebUI go to **Settings → Connections** and add a connection:
+   - **API Base URL**: `http://localhost:1416` (or `http://hayhooks:1416/v1` when using Docker Compose)
+   - **API Key**: any value (Hayhooks does not require auth)
+
+3. In a new chat, select your deployed pipeline as the model.
+
+Pipeline wrappers must support chat completion (e.g. implement `run_chat_completion` or `run_chat_completion_async`). See [OpenAI compatibility](https://deepset-ai.github.io/hayhooks/features/openai-compatibility) and the [pipeline examples](https://deepset-ai.github.io/hayhooks/examples/overview/) for implementation details.
+
+### Optional: Open WebUI events
+
+For status updates, notifications, and tool-call feedback in the UI, use helpers from `hayhooks.open_webui` in your pipeline and stream `OpenWebUIEvent` objects. See [Open WebUI Events](https://deepset-ai.github.io/hayhooks/examples/openwebui-events/) and the [open_webui_agent_events](https://github.com/deepset-ai/hayhooks/tree/main/examples/pipeline_wrappers/open_webui_agent_events) and [open_webui_agent_on_tool_calls](https://github.com/deepset-ai/hayhooks/tree/main/examples/pipeline_wrappers/open_webui_agent_on_tool_calls) examples.
+
+## License
+
+`hayhooks` is distributed under the terms of the [Apache-2.0](https://spdx.org/licenses/Apache-2.0.html) license. Open WebUI is subject to its [own license](https://github.com/open-webui/open-webui).