# Tasks
Stop juggling spreadsheets, Slack threads, and scattered task boards. Valdr Tasks give you **one place to see what's happening, what's blocked, and what's done**—with AI agents ready to pick up work the moment you assign it.

**Tasks are the unit of execution in Valdr.** Each task carries its spec context, ownership, status, reviews, and agent session history—so work stays explainable even weeks later. When someone asks "what happened on this?" the answer is one click away.

Every task connects to its plan, links to requirements, and provides a launchpad for AI agents. That means traceable, reviewable, auditable delivery that scales with your workload—not against it.

{{< thumbcard src="/images/ui/tasks/task-list-overview.png" alt="Tasks overview showing status tabs and task table" caption="See all work across projects at a glance—filter by status, assignee, or project in seconds" >}}

## What this section covers

- **[Task List](task-list/)** — Filter, sort, and triage tasks across all projects
- **[Task Detail](task-detail/)** — Deep dive into individual tasks with tabs for description, comments, reviews, sessions, and activity
- **[Sidebar & Metadata](sidebar/)** — Quick access to status, ownership, and task properties
- **[Launching Agents](agent-launch/)** — Spin up AI agents directly from tasks with full context

## Why Valdr tasks are different

Traditional task trackers show you *what* to do. Valdr tasks show you *what*, *why*, and *how it connects* to your broader delivery goals—then let you hand work off to an AI agent with one click.

**Spec-driven context** — Every task links to plan requirements. No more "what were the acceptance criteria again?" moments. The spec is right there, always.

**Agent-ready by default** — Click **Launch**, and the agent inherits the task context: description, linked requirements, acceptance criteria, checklist state, and recent activity. No copy-paste prompt stitching. No re-explaining what you need.

**Review-gated quality** — Built-in review workflows make quality enforceable. Request reviews from humans or agents, get structured feedback with severity ratings and specific file references—not vague "looks good" approvals.

**Full audit trail** — Status transitions, assignee changes, comments, review decisions, checklist updates, and complete agent session history—all logged automatically. When you need to answer "what happened?", the receipts are there.

## Task lifecycle

Tasks flow through six statuses that match how real teams actually work:

| Status | What it means |
|--------|---------------|
| **Backlog** | Captured but not yet prioritized |
| **To Do** | Ready for work this sprint |
| **In Progress** | Actively being worked on |
| **In Review** | Submitted for review; awaiting approval |
| **Verified** | Review passed; ready for final sign-off |
| **Done** | Complete and closed |

Watch your work flow from left to right. When something stalls, you'll see it immediately.

## Get started in 60 seconds

{{% steps %}}

### Navigate to Tasks

Click **Tasks** in the left sidebar. You'll see all tasks across your projects, with status counts at the top.

### Filter to what matters

Click a status tab to focus on specific workflow stages. Use the project dropdown to scope to one project. Search by task key, title, or assignee.

### Open a task

Click any row to see the full detail view—description, checklists, review history, and agent sessions all in one place.

### Launch an agent

From the task sidebar, select a launcher preset and click **Launch**. An AI agent starts working immediately, inheriting the complete task context. Track progress in **Agent Sessions** and **Activity**, or work on something else and check back later.

{{% /steps %}}

## Common outcomes

Developers using Valdr tasks consistently see:

- **Shorter standups** — When task context includes requirements, acceptance criteria, and session history, there's less to explain verbally
- **Faster review cycles** — Structured feedback with severity and file references makes fixes obvious; no back-and-forth clarification
- **Answerable "what happened?"** — Agent sessions and activity logs mean every question has a traceable answer

## Next steps

{{< cards >}}
  {{< card link="task-list/" title="Task List" icon="table" >}}
  {{< card link="task-detail/" title="Task Detail" icon="document-text" >}}
  {{< card link="sidebar/" title="Sidebar & Metadata" icon="adjustments" >}}
  {{< card link="agent-launch/" title="Launching Agents" icon="play" >}}
  {{< card link="../agent-sessions/" title="Agent Sessions" icon="terminal" >}}
{{< /cards >}}

**Ready to see it in action?** Open your Tasks dashboard and launch your first agent—most developers see results in under 2 minutes.