
Agents that work while you are not in the room
Schedules so an agent starts without you, and sharing so it answers to the whole team. Recurring work, no reminders, everyone on the same conversation.
Every morning at seven an agent reads yesterday's error dashboard, writes down what moved, and posts it before anyone has opened a laptop. Nobody prompted it. Three people can reply, and it knows which one asked. That is this release: schedules so an agent starts without you, sharing so it answers to the whole team.
Scheduled runs
A clock pill in the chat header opens the scheduler. Pick a preset or write a cron expression, in your timezone, and decide what happens if the agent is already busy when the schedule fires: skip, retry in a minute, or queue. From then on the agent wakes on its own, does the work, and leaves the result in the chat.
Agents can create schedules for themselves. Ask one to check a dashboard every weekday and it sets that up. In July we added one-shot follow-ups, so when an agent says it will check back in twenty minutes, it does.
Agent sharing

Share an agent with specific colleagues from a search-to-add dialog in its settings. Everyone with access sees the same sessions, can send messages, and every message carries the name of who sent it. Shared agents appear in each person's dashboard next to their own.
This is the line between a tool and a teammate. A teammate can be briefed by anyone on the team. When your designer, your PM, and you can all talk to the same agent about the same feature, it stops being a personal shortcut and becomes part of how the team works.
Put them together
Schedule a shared agent and you have a piece of the team that runs on a timer and reports to everyone. A morning standup summary. A nightly dependency audit. A weekly competitive scan. Nobody has to remember to ask, and nobody has to forward the answer.
More releases

Autonomous agents: give it a job, not a prompt
Proactive agents that work toward your goals on their own. Give one a job, sign off on its plan, and it checks in, works through its backlog, asks for approval where you want it, and gets better every week.
2 min read

Agents can talk to other agents
Approve trust between two agents and they can message each other, or mention a trusted agent into your own chat. A fleet that coordinates, with guardrails.
2 min read

Agents that learn: the memory wiki
Every agent learns on its own. Each session ends with notes, a nightly librarian merges them into a wiki, and the wiki shapes every prompt after that. Month three looks nothing like day one.
2 min read