Projects & Tasks

Project management, with agents on the team.

Boards, lists, saved views, milestones, and request intake: everything you expect from a mature tracker, without the migration regrets. Assign any task to a person, or to a governed AI agent.

Agents as assignees

Assign it to a person. Or to an agent.

Any shared agent can be a task assignee. On a coding task, Start opens Start coding work: leave a note for the agent, pick the coding agent and the mode (Agent, Plan, or Ask), and it opens a live sandbox session. Progress posts back to the task discussion, and you steer from there: reply "retry it" or "also update the README" and the agent passes it on.

When its pull request opens, the task moves to In review by itself, and the pull request description carries the agent's confidence. Project-level agent guidance tells agents how your team works.

  • My tasks: one cross-project board for everything assigned to you
  • Workload views include agent assignees
  • Work threads keep discussion attached to each item

Connected to code

Tasks that know about your repos

Link GitHub, GitLab.com, or Azure Repos repositories to a project: mirror GitHub and GitLab issues as tasks or requests, connect pull requests to tasks, and start coding sessions straight from a task. Everything stays linked: task, session, PR, preview. A task can close itself when its pull requests merge, and pipeline deployments run from the project when it is time to release.

AI review and request refinement come from the AI delivery templates: Code workflows a team creates for the projects that need them. The project's Delivery card counts merged pull requests, the agent merge rate, and time to merge, and project managers get a development status view of every piece of work in progress, no git knowledge needed.

When the assignee is an agent, the whole loop runs inside the same card.

Story

See how every piece of work travelled

Story follows work from the chat message or support ticket that started it, to the request, the task and its sub-tasks, the workflow runs, the coding session, and the pull request with its reviews. Activities lists what people and agents did across all of them, newest first. The story map draws the same items as cards in the order the work travelled, and keeps up as it happens.

Stories lists every story that is moving right now and every one that finished. Volobox builds them from recorded activity and links: no AI writes or summarizes a story.

Overview and sharing

One picture of the project, for the team and the client

The fundamentals

And yes, it's a serious tracker on its own.

No feature checklist surprises when you move over.

Boards & views

Kanban and list views with saved filters per project and per person, and bulk edit for many tasks at once.

Requests

Intake for bugs and ideas, triaged one by one or in bulk. Tasks handle them: create one task for several requests or add them to an existing task, and each request closes when its task completes.

Milestones

Time buckets that group tasks and anchor planning.

Labels & areas

Tags for filtering, areas for ownership and request routing.

Members & roles

Viewer, Collaborator, Manager, Admin, per project.

Activity

A project-wide feed of what changed, by whom: humans and agents alike. Each task and request keeps its own field-by-field history in its Story.

Bring your board over, and give it new teammates.