# Task Detail
Open any task and you get the complete picture: what needs to happen, who's working on it, what the agents have done, and every decision along the way. No more hunting through Slack threads or email chains. **One URL, complete context.**

{{< thumbcard src="/images/ui/tasks/task-detail-overview.png" alt="Task detail page showing Overview tab" caption="Description, acceptance criteria, requirements, and checklists—all visible without scrolling" >}}

## What this page covers

- Task header and navigation
- Six detail tabs: Overview, Comments, Reviews, Agent Sessions, Activity, Metadata
- Working with checklists
- Editing task descriptions

## Task header: instant orientation

Every task detail page starts with the essentials:

- **Breadcrumb** — One click back to the task list
- **Task key** — Copy-pasteable identifier (`TESTP-22`)
- **Status badge** — Current workflow stage, color-coded
- **Project name** — Click to see related tasks
- **Title** — What this task accomplishes

You always know where you are and how to get back.

## Overview tab: the source of truth

This is where you'll spend most of your time. Everything a human or agent needs to execute the task lives here.

### Description

Rendered Markdown with structure that matters:

- **Summary** — What needs to be done (the "what")
- **Acceptance Criteria** — Specific conditions for completion (the "how we know it's done")
- **Linked Requirements** — References like `REQ-1`, `REQ-2` connecting to the plan (the "why")

Click the **Edit** button to update. Changes are versioned in the Activity tab.

### Checklists: your definition of done

Interactive checkboxes that track completion criteria:

- Click any item to toggle—green checkmark means done
- Progress persists across sessions
- Checklist updates trigger Activity events

> [!TIP]
> **Ship with confidence:** When all checklist items are green, you know the task meets its acceptance criteria. No guessing, no "I think it's done."

## Comments tab: async collaboration that works

{{< thumbcard src="/images/ui/tasks/task-comments-tab.png" alt="Comments tab with discussion input" caption="Threaded discussions stay with the task—not lost in Slack" >}}

Stop losing context in chat apps. Task comments stay **attached to the work**:

- Markdown formatting for code blocks, links, lists
- Timestamps show when each comment was posted
- Both humans and agents can participate
- Full history preserved for the lifetime of the task

Use comments for:
- Clarifying requirements
- Documenting implementation decisions
- Agent-to-human handoffs
- Capturing "why did we do it this way?" context

## Reviews tab: quality gates that catch real issues

{{< thumbcard src="/images/ui/tasks/task-reviews-tab.png" alt="Reviews tab showing structured feedback" caption="Structured review feedback with severity, location, and specific remediation steps" >}}

This is where Valdr's review workflow shines. No more "LGTM" drive-bys.

### Requesting a review

1. Enter the reviewer handle (`@documentation-reviewer`)
2. Add optional context for the reviewer
3. Click **Request**

The reviewer—human or agent—gets notified with full task context.

### What reviewers produce

Structured feedback that's actually actionable:

- **Summary** — High-level assessment
- **Issues** — Specific problems with severity (major/minor), file location, line numbers, and remediation steps
- **Validation** — What was tested, what commands ran
- **Recommendation** — Ready for sign-off, needs revision, or reject

### Review statuses

- **Pending** — Awaiting review
- **Changes Requested** — Needs work before approval
- **Approved** — Good to merge/complete

Plus a **score** (0-100) for quantitative tracking across your work.

> [!NOTE]
> **Why this matters:** Reviewer agents produce this structured output automatically. Every review is thorough, consistent, and auditable.

## Agent Sessions tab: see what your agents actually did

{{< thumbcard src="/images/ui/tasks/task-agent-sessions-tab.png" alt="Agent Sessions tab with session history and launcher" caption="7 sessions, full token usage, and the ability to launch more—all from one tab" >}}

This tab answers: "What have agents done on this task?"

### Session history

Every agent run is logged:

- **Session ID** — Click to view the full session transcript
- **Role** — task, reviewer, or auditor
- **Status** — Running or Closed
- **Provider** — Which AI backend (Codex, Ollama, Anthropic)
- **Duration** — How long it ran
- **Time** — When it happened

### Session comparison

The analytics panel shows patterns across sessions:

- **Tokens** — Input, output, cached (spot cost drivers)
- **Tools** — Tool call frequency
- **Commands** — Shell commands executed
- **Duration** — Time comparison

Group by Provider, Model, or Role to understand performance differences.

### Agent launcher

Launch new sessions without leaving the task. Full configuration available—see [Launching Agents](../agent-launch/) for details.

## Activity tab: complete local audit trail

While **Agent Sessions** shows what automation ran, **Activity** captures every state change—from humans and agents alike.

{{< thumbcard src="/images/ui/tasks/task-activity-tab.png" alt="Activity timeline showing events" caption="Every change logged—who did what, when, and in what order" >}}

When someone asks "what happened on this task?", you have the answer:

- **Checklist events** — Items checked/unchecked
- **Comments** — Added or deleted
- **Status changes** — Who moved it, when
- **Agent sessions** — Started, completed

Filter by event type. See exact timestamps. **Full accountability, zero extra work.**

> [!NOTE]
> **Data retention (local-first):** Valdr stores tasks, reviews, sessions, and activity locally on your machine. History is retained as long as the underlying items exist. You can delete projects, plans, tasks, or sessions at any time to reduce workspace bloat or reset context.

## Metadata tab: all properties in one view

{{< thumbcard src="/images/ui/tasks/task-metadata-tab.png" alt="Metadata tab showing task properties" caption="Quick access to all task properties—assignee, status, tags, and checklists" >}}

When you need to update task properties quickly:

- **Assignee** — Reassign with one click
- **Reporter** — Who created this
- **Points** — Story point estimate
- **Priority** — Urgency level
- **Status** — Workflow stage
- **Tags** — Categorization labels
- **Links** — External resources (PRs, docs, designs)
- **Checklists** — Same interactive lists as Overview

Everything in one view. No hunting through menus.

## The result: tasks that actually track work

With Valdr task details, you get:

- **Complete context** — No "let me find that Slack message"
- **Structured reviews** — No "LGTM" without substance
- **Agent visibility** — Know exactly what automation did
- **Complete local audit trail** — Every change logged for as long as the task exists

This is the difference between "task management" and **task operations**.

## Next steps

- **[Sidebar & Metadata](../sidebar/)** — Quick actions from the sidebar
- **[Launching Agents](../agent-launch/)** — Spin up agents from any task

**Open a task now** and explore the tabs. See how much context is already there—imagine never losing track of work again.