👀 Getting an agent to simply show "I'm on it" can change how much users trust it. LangChain just shipped a small but effective feature for exactly that. Title: Build great out-of-the-box user experiences with Managed Deep Agents URL: Managed Deep Agents v0.9 adds a new reactions API for Slack channels, letting an agent post an emoji acknowledgment before it even replies. Here are 3 highlights. 👀 Simple emoji reactions Pass the reactions argument either a static emoji string or an async callable that takes context and returns one. You can start with a lightweight if/else, like returning 🐛 when the message contains "broken" and 👀 otherwise. 🧠 Dynamic picks via the Jev classifier With TypeSafeClassifier(model="typesafe/jev-latest"), the message gets auto-classified into categories like bug, incident, review request, or waiting, then mapped to the best emoji. If confidence drops below 0.25, it falls back to a safe 👀. 🎭 Skeuomorphic UI with custom emoji Custom Slack emoji shortcodes are supported too, so teams can use branded reactions to visually signal that an agent has just picked up a task. The takeaway: unglamorous UX details like loading states are often what decide whether users actually trust and keep using an agent. #LangChain# #AIAgents#