<?xml version="1.0" encoding="UTF-8"?><oembed><type>video</type><version>1.0</version><html>&lt;iframe src=&quot;https://www.loom.com/embed/a88d525111b1445cb90db342ade09ebc&quot; frameborder=&quot;0&quot; width=&quot;1670&quot; height=&quot;1252&quot; webkitallowfullscreen mozallowfullscreen allowfullscreen&gt;&lt;/iframe&gt;</html><height>1252</height><width>1670</width><provider_name>Loom</provider_name><provider_url>https://www.loom.com</provider_url><thumbnail_height>1252</thumbnail_height><thumbnail_width>1670</thumbnail_width><thumbnail_url>https://cdn.loom.com/sessions/thumbnails/a88d525111b1445cb90db342ade09ebc-56b0ee8adfa50951.gif</thumbnail_url><duration>178.39</duration><title>LiteLLM Agent Platform Demo Video - Claude Code, Codex</title><description>This Loom demonstrates how to use the agent platform to run Cloud Code in sandbox environments with template-based repositories and environment variables. It shows creating a new repository template in the UI by providing values like a LIDL repo URL and an OpenAI API key, then creating an agent assigned to that template with a selected model, optional PR standards skill, and optional MCPs. The developer then uses the Light LME Agent Platform CLI to choose the Cloud Code 2 agent, which provisions a sandbox preloaded with the repo and uses a vault proxy to place credentials on outgoing requests. The demo includes SSHing into the sandbox to confirm the presence of the Light LLM repo files and receiving responses from CloudNow.</description></oembed>