Most teams start the day by piecing together what happened overnight: checking Slack, scanning JIRA, chasing down updates. In an AI native organization, that work happens before anyone opens a laptop. This is how our team runs its morning reports with Claude — not as a demo of what AI could do someday, but as the routine that already gets us through our standups.
The result is simple to show. Each morning, a generated report is waiting on screen with everything a person needs to know: what moved, what needs their attention, what needs their approval, and who they should reach out to.
Shared team skills: roles and daily reports
Everything starts with Claude skills that we distribute to everybody on the team. Once the team is configured, everyone gets the same setup, and we can keep improving the skills we run together.
A few skills do the heavy lifting:
- Roles — a role is just something you run: you talk about what you do, and it produces a living document so the AI knows what you're responsible for.
- The evening report — when you're done for the day, you type the command and it scoops up everything you did, makes a report, and dumps it into S3.
- The morning skill — it takes everybody's evening reports and rolls them up into a personalized report of what you need to know.
Because the reports land in S3, the agents can see everything. That shared data layer is what makes coordination possible at all.
How the coordination actually works
Right now the team runs the role skill, but the plan is to pull in absolutely everything: all the meetings, all the JIRA tickets. Every event gets summarized, and each person gets a report built around what matters to them specifically.
So if somebody is working on something that may need your approval, the AI agent coordinates that to you — you need to know this, you need to reach out to this person. And it comes with recommendations: start this task, this person is blocked, book this meeting. The agent can handle all of that follow-through itself.
We're basically automating our decision-making as an organization.
Always-on agents in the cloud
The other half of the setup is Claude running on our cloud environment. Everybody can work through the remote environment, and because the agents run there, you can close your laptop and they keep going. Our agents run 24/7 — like a teammate that never clocks out.
Combined with the shared skills and the S3 data layer, that means the reports keep building themselves whether anyone is watching.
Turning the report into a habit
The workflow only pays off because it's a habit: run the report, read it, bring it to the meeting. That's the daily loop. The reports also go out to Slack, so updates reach the team where the team already is.
These are some of the improvements we've made to how we work as an AI native organization — everything going into S3 so the agents can see everything, and the coordination and decisions increasingly handled by the agents themselves. That's what it's like working in an AI native organization.
Transcript
So this is how my team does their morning reports and communicates with claude. So in an AI native organization, this is how you can move a little bit faster. So we have CLAUDE skills that we distribute everybody to. The team gets it. So our cloth's configured, everybody gets this stuff. So we can improve the skills uh, that we're uh, running now what we can do is we have these files or these skills that save data to S3. So a role, for example, is just something you run. You talk about what you do. And this is a living document just so the AI knows um, what you're responsible for. Um, then we have an evening report tool. So what this does right here, this skill, when you type, uh, evening, uh, when you're all done, you basically just, it scoops up everything that you've done, it makes a report and it dumps it into S3. Um, the morning skill is it takes everybody's skills or everybody evenings report that are done and then it scoops them up into a personalized report for you of what you need to know. So right now we're doing, um, we're having the role, right, um, but then eventually we're pulling in, uh, we pull in all the meetings, we pull in all the JIRA tickets, we pull in absolutely everything. So you get a report of what you need to know. If somebody's working on something that may need your approval or whatever. The AI agent is going to be able to coordinate um, to you of like, hey, you need to know this, you need to reach out to this person and then have a bunch of recommends of, oh, hey, start this task, this person is blocked, book this report and the agent can handle all that stuff. So we're basically automating all our decision making and um, uh, in the organization. And so this is with a cloud integration. We also have CLAUDE integrated. We have um. Oh, just a minute here. Oh, and these, these also go out to slack is what they. Oh yeah. So yeah, we have. I'll just move this down, uh, down here we have remote env. And so now everybody can basically we're running CLAUDE on our cloud. So you close your laptop, it's still running. So our, our agents can run 24 7. So these are some improvements we, we've made to our AI, like as an AI native organization where everything's going into S3 so the agents can see everything and then we can start automating our decisions. And so this is what it's like working in an AI native organization.
Frequently Asked Questions
An evening report skill scoops up everything you did during the day, makes a report, and dumps it into S3. The morning skill then takes everybody's evening reports and rolls them into a personalized report of what you need to know.
A role is something you run where you talk about what you do. It produces a living document so the AI knows what you're responsible for.
Claude runs in the team's cloud environment, so when you close your laptop it is still running and the agents can work 24/7.
Yes — the reports also go out to Slack, so updates reach the whole team there.
The goal is to pull in all the meetings and all the JIRA tickets, so every event is summarized and each person's report includes approvals, blocked teammates and recommended next actions.