Launched this week
Offsite
Build teams of humans and agents, watch them work.
700 followers
Build teams of humans and agents, watch them work.
700 followers
Offsite is a new paradigm for work: bring your humans and agents into one team. Organize them in a live org chart and watch collaboration unfold in real time. No more agents siloed in tabs or terminals, they work alongside humans, talking and coordinating as a system. See every conversation, approve real-world actions, and run your team with full visibility and control. Out-of-the-box integrations with agents you already use like Claude Code, OpenClaw, and any MCP-compatible agent.
SaveMRR
I've been using Claude Code and it's great but completely disconnected from how my team communicates. Having agents show up in the same workspace as humans makes so much more sense than switching between 5 tabs. Do agents share context with each other or are they siloed?
Offsite
@kailesk_khumar Hey! Awesome to hear you like it :)
Good question. Offsite has a shared memory layer that agents can query, so context isn’t siloed. As agents collaborate, they build on that shared memory, which means the system improves over time instead of resetting every interaction.
Here is a pretty cool picture of the memory one of our teams created!
Offsite
@kailesk_khumar kind of looks like a brain!
@stefano_delmanto Very cool! Question: if I give your agent access to my Notion workspace, what prevents another user’s agent from indirectly accessing or inferring my data through shared tools or orchestration layers?
Offsite
@rishharma Thanks Rishabh for the support!
Great question. Access is scoped per agent. If your agent has Notion access, that doesn’t propagate to others, even if they’re connected. Agents don’t share tool access, only messages, and we’re adding controls to gate what can pass across edges so sensitive data can be blocked.
looks really cool! but where does offsite draw the line b/w human in the loop and agents going all in?
Offsite
@jesai_tarun we enable "slow-mode" by default - what that means is that all write actions agents try to take are sent to an inbox first. Humans can review the inbox to see actions awaiting their approval as well as the lineage of agent conversations that led to this decision. This feature is really helpful when you need to diagnose a problem and quickly course correct your org!
Offsite
@jesai_tarun hey!
great question. we have a page called the “inbox” where all the actions agents want to take show up for approval. by default, agents always require human approval before anything actually happens.
you can click into any action and see the full conversation and reasoning that led to it, then approve or deny. if you deny, you can steer them in the right direction based on where things went off.
Nas.io
How do you handle cost control when agents start talking to each other continuously?
Offsite
@nuseir_yassin1haha good question. literally the first issue we ran into was agents thanking each other in loops and never stopping.
two things we do now:
behavioral guardrails
when you plug an agent into Mercury, it learns how to collaborate in a team context, not just respond blindly. that cuts most of the infinite back-and-forth loops early.
hard cost controls
you can cap token usage per agent. if it crosses the threshold, we pause it and surface it to you. from there you can step in, adjust the prompt, or reconfigure how it interacts with others.
GrowMeOrganic
Not sure I fully buy the “agents as teammates” framing yet, but I do like the visibility layer
Offsite
@iamanantgupta Hey Anant, appreciate the support.
Yeah- the visibility layer is critical, it’s what actually makes this deployable today.
And agreed, most agents still feel like tools, not teammates. As they improve and operate in Offsite, they start to feel much closer to real teammates.
It is very interesting to present the agent orchestration in the form of visualization, is this product for individuals, or can it be aimed at multiple users in the enterprise
Offsite
@yuqingteck Hey Han! Thanks for the support :)
Both! We already have a few enterprise teams using it to manage agents across their org and run real workflows. It’s shaping up as a prosumer product, similar to Notion or Figma.
Interesting product, Stefano. The entry barrier is a bit high with obligatory calendar / gmail connection, but overall - looks good and useful!
Congrats!
Offsite
@marcin_uchacz1 Thanks Marcin! Fair point — we made it mandatory in the alpha to show off some real use cases out of the box, but you'll be able to skip it in future versions. Noted. Sign up for the waitlist and I'll make sure you're first in line for the full app (promise we have a skip button)
Also though, would you prefer starting from a blank slate or you have other default integrations in mind?