Create an agent for one specific task.
Describe the goal, choose the sources and tools, then test the result before sharing the agent with your team. Each agent is packaged as an application, with its triggers, approvals and budget, and versioned like software.
88%Installer →
71%Installer →What you get with Agents & Applications.
Everything below is in the app today.
Agent Builder
Describe the goal, the trigger, the actions and the expected result. The Builder composes skills, detects tools and plans tasks.
Reusable skills
Packaged know-how: instructions, tools, output schema. Installable on any agent.
Triggers
Chat, Slack command, cron, webhook, API. One agent, several entry points.
Approvals, SLA, budget
Approval rules, time targets with on-call, daily spend cap, active hours.
Factory
A catalog of ready-to-install applications, skills and tools, published to your organisation or your space.
Versions and rollback
Each application is described in YAML and deployed from the command line, versioned like software: publish, pin, roll back. Every run keeps the agent version that produced it.
In three steps.
You describe the work, you set the boundaries, you supervise the results.
- 01
Describe
Goal, trigger, actions, result. The Builder proposes a first version.
- 02
Test
Chat with the agent, invoke its tools in a sandbox, adjust the instructions.
- 03
Deploy
Publish a version, wire a trigger, set approvals and budget.
Let’s discuss your first use case.
Describe the work you need done and the tools you use. We’ll prepare the demonstration around your needs.
Talk with someone who knows the product and your project.
Bring a recurring task and the result you need.
We’ll explain the next steps and the conditions for beta access.