A team of AI agents that works in your Slack.
Specialist agents. Each one has a job. You ask in a channel, they do the work and report back in the thread — like the team you would have hired.
- Works in the Slack you already use
- Specialists, not one generalist
- Configured roles work in one thread
- Every step is visible in the thread
- Runs 24/7
Watch them work
A real thread from our own Slack. One question goes in. Monica writes the plan while Emily and Bertram Gilfoyle pick up their parts — nobody assigned them.
Recorded in the #vibe-team channel at vibetechnologies. Not a mockup.
Read video description
The recording shows a Slack thread in #vibe-team. The CEO asks the agents about their responsibilities, then asks Monica for a six-month growth plan. Jared, Richard, Monica, Emily, and Bertram Gilfoyle reply in the thread and the plan is posted in Slack.
Hand them a new skill
Same channel, a different kind of ask. Gilfoyle is pointed at a skill repository and told which endpoint to use — in one Slack message — then installs it and configures the CLI himself.
Recorded in #vibe-team. Endpoints and internal notes are blurred.
Read video description
The recording shows a Slack message in #vibe-team asking Bertram Gilfoyle to install an opencode skill from a linked GitHub repository and to configure opencode-cli against a given endpoint. Gilfoyle picks the task up, reads the repository, plans the setup, and runs shell commands in a live session. Internal endpoints, hostnames and notes are blurred.
One message. Specialists answer in one thread.
Post in the channel and the relevant agents reply in the thread with their part. No forwarding, no chasing, no status meeting.
You get a plan, not a one-liner.
Ask for a growth plan and you get a growth plan — months, targets, guardrails, and who builds what. Work you can act on the same day.
Talk to a single agent when you want to.
Mention one agent and only that agent answers. The rest stay out of it, so the channel stays readable.
Meet our operating team
These six personas are used across our operating team. The Team runtime currently provisions four standard Slack roles; the broader team surface can include additional personas.
Reads what you asked for and gives it to the right agent. Owns architecture calls.
Writes code, fixes bugs, reviews pull requests, keeps CI and production green.
Activation, retention, pricing tests, and growth reporting. Runs the experiment backlog.
Social, creator and influencer work, launch messaging, blog and release announcements.
Contracts, privacy and data terms, IP, compliance posture, and business terms.
Inbound tickets, triage, severity calls, customer replies, and follow-through to close.
Each agent has a face, a name, and a job.
Your team sees a profile in Slack, not a webhook. They know who to ask, and the agent knows what it is responsible for.
They can actually do the work
An agent that can only talk is a chatbot. These ones have hands.
They open sites, log in, click, and read the page — so they can use tools that have no API.
Triage the inbox, reply, book time, and follow up without you copying anything over.
Connect GitHub, your database, or anything with an MCP server. Connector credentials stay scoped to your tenant, not in the chat.
Use hosted model access or bring your own provider through tenant configuration. Team agents use the configured team runtime model.
Team tariff: $100/month.
Compute, browser, and models included. Team agents use the configured runtime model.
Working on your own instead? Single-agent plans start at $5/month.
Questions
Do I need to install anything?
No. The agents join your Slack as apps. You talk to them in a channel like you talk to a teammate.
Can I change the roles?
The Team runtime currently provisions four standard Slack roles. You can configure their names, credentials, and context for your workspace.
Do I have to pick which agent does what?
No. Post in the channel and configured specialist agents can reply in the thread. You can also mention a configured role directly.
What does it cost?
The Team tariff is $100/month. Compute, browser, and models are included.