AI agents
Main working loop
DODVIR
CLI works with a local repository, SOLO runs one personal workstream, and MULTIAGENT coordinates several roles around one result.
Choose the setup that fits your work. Models and provider keys stay with you; approved access and updates are managed in your account.
AI agents
Main working loop
Working cycle
The task moves to the finish
Roles
Each role has its own area of responsibility
CLI works with a local repository, SOLO runs one personal workstream, and MULTIAGENT coordinates several roles around one result.
CLI
Run DODVIR from the terminal in a local repository and keep every change reviewable.
OpenSOLO
Give one personal agent workstream a task and receive a reviewable result and next step.
OpenMULTIAGENT
Coordinate several agent roles around a shared task, context, handoffs, and final review.
OpenSolo is tuned to your role:
A simple loop you control: you bring tasks, agents do the work, you accept the results.
Step 1
We connect the repository or process and tasks land in the backlog. The agent studies the project and keeps its own knowledge base about it.
Step 2
Plan -> implement -> test -> review. When something is unclear, the agent asks a question instead of guessing.
Step 3
Small changes, automated checks, and clear merge rules. You decide what ships to the product.
Choose a product and tell us about your task. After review, your account shows what is available to you:
Choose the format and leave a contact. We will define the right configuration.