Add visualizer docs

docs/mkdocs.yml

@@ -274,6 +274,10 @@ nav:
         - Social Media:
           - Twitter: "swarms_tools/twitter.md"
 
+    - Swarms Utilities:
+      - Overview: "swarms/utils/index.md"
+      - Visualizer: "swarms/utils/visualizer.md"
+
     - Swarms Memory:
       - Overview: "swarms_memory/index.md"
       - Memory Systems:

docs/swarms/utils/index.md

@@ -0,0 +1,7 @@
+# Swarms Utilities
+
+The `swarms.utils` package provides small helper modules that simplify common tasks throughout the framework. These utilities range from data formatting helpers to system monitoring tools.
+
+## Available Guides
+
+- [Visualizer](visualizer.md) – real-time dashboard for agent output and resource usage.

docs/swarms/utils/visualizer.md

@@ -0,0 +1,42 @@
+# Visualizer Utility
+
+`SwarmVisualizationRich` is a live dashboard built with Rich that lets you watch agents work while monitoring system resources.
+
+## Starting the Visualizer
+
+```python
+from swarms.utils.visualizer import SwarmVisualizationRich, SwarmMetadata
+from swarms import Agent
+
+# create agents
+agent = Agent(name="DemoAgent")
+
+metadata = SwarmMetadata(name="Demo Swarm", description="Example swarm")
+
+viz = SwarmVisualizationRich(swarm_metadata=metadata, agents=[agent])
+
+import asyncio
+asyncio.run(viz.start())
+```
+
+The `start()` method runs an asynchronous event loop that continually refreshes the console layout.
+
+## CPU and GPU Metrics
+
+When `update_resources=True`, the visualizer displays current CPU core count, memory usage, and GPU utilization. GPU statistics rely on `pynvml`; if no GPU is found the panel shows `No GPU detected`.
+
+- **CPU** – logical core count returned by `psutil.cpu_count()`.
+- **Memory** – used/total RAM with percentage from `psutil.virtual_memory()`.
+- **GPU** – per‑device memory consumption via NVIDIA's NVML.
+
+## Embedding in Custom Workflows
+
+You can stream agent output into the visualizer from any workflow:
+
+```python
+viz.log_agent_output(agent, "Task started...")
+result = await some_coroutine()
+viz.log_agent_output(agent, f"Result: {result}")
+```
+
+Combine this with your swarms or structured workflows to obtain a live view of progress and system health.