Task Tracker — User Guide | YoBench

What the Tracker module does

The module aggregates tasks from every corporate issue tracker into a single place. Instead of switching between eight browser tabs (Yandex Tracker, JIRA, GitHub, GitLab, Linear, Asana, Trello, Redmine), you see one unified list or kanban board for tasks across all projects and teams — with filters, search, AI estimation and a local cache.

What you get:

• One queue from 8 systems — every integration stores data in the encrypted local database, so you can keep working offline.
• List and Kanban views — switch between a flat list and a status-based board with one click.
• AI complexity estimation — a button right inside the task detail: AI reads the description, comments, and checklist, and returns hour estimates with rationale.
• Rich actions on native tasks — for internal tasks: status changes, comments (rich text), time tracking, checklists, labels, task links and custom fields.
• Custom fields with full-text search — define text, number, date or select fields per project; search across them as fast as core fields.
• Subtabs in the task detail — Description, Checklists, Comments, Activity, Links and AI estimate live on dedicated subtabs so the panel stays readable on long tasks.
• Per-connection scheduler — each tracker connection has its own poll interval (poll_minutes), so a fast-moving JIRA can refresh every 2 minutes while a calmer Redmine polls hourly.
• Paginated comments and infinite scroll — long threads load in pages, large task lists keep scrolling smoothly without a stall.
• Read-only view of external tasks — tasks from connected trackers appear inside YoBench; edits happen in the source system.
• Background sync — every connection is polled on its own interval, no manual intervention required.

Supported providers and connection parameters

Eight adapters are implemented in code and available from the "Add connection" form.

Yandex Tracker

• Org type (cloud or 360)
• Org ID
• OAuth token
• Optional: assignee filter, custom query

JIRA Cloud

• Host — e.g. https://yourteam.atlassian.net
• Email
• API token
• Optional: project keys (for filtering)

GitHub Issues

• Personal access token
• Repos in owner/repo form (required)
• Optional: assignee filter

GitLab Issues

• Personal access token
• Project IDs or paths (required)
• Optional: GitLab URL (for self-hosted)

Linear

• API key
• Optional: team keys, assignee emails

Asana

• Personal access token
• Project GIDs (required)
• Optional: assignee names

Trello

• API key
• API token
• Board IDs (required)

Redmine

• Redmine URL
• API key
• Optional: project identifiers

UI capabilities

List view

Table columns:

Kanban board

Columns correspond to the statuses of the selected project or status mapping. Cards drag between columns — for native tasks this changes the status; for external tasks it syncs back to the source tracker.

Filters

• By status, priority, assignee
• By project, label, link type (internal / external)
• Search by title
• Custom filters on task fields
• Sort by: updated, priority, created, due date

Actions on tasks

Native tasks (created inside YoBench):

• Create / edit / archive
• Status change (drag-and-drop on the board or via the detail panel)
• Comments in a rich text editor (TipTap), paginated with infinite scroll on long threads
• Time tracking (estimate_hours, spent_hours)
• Checklists with items
• Labels (project-wide registry)
• Task links: blocks / depends on / duplicate
• Custom fields: text, number, date, select — indexed for full-text search alongside title and description
• Move between projects

External tasks (from connected trackers): view-only plus AI estimation. Make edits in the source system — they sync back on the next poll.

Custom fields and FTS

Per project you can define custom fields of type text, number, date or select. Once defined they appear in the task editor and as filterable columns in the list view. The full-text index covers core fields (title, description, comments) and every custom text field — so searching for a customer code, ticket reference or release tag finds tasks no matter where the value lives.

Task detail subtabs

The right-side detail panel keeps the header (status, priority, assignee, due date) always visible and splits the body into subtabs:

• Description — task body and metadata.
• Checklists — multiple named lists with drag reordering.
• Comments — paginated thread with rich text and replies.
• Activity — chronological feed of all changes.
• Links — blocks / depends-on / duplicates.
• AI estimate — run, view and re-run AI complexity estimates.

AI task estimation

The task detail panel has an AI estimation tab. Clicking Run AI estimation makes the module:

1. Collect the task title, description, comments and checklist items.
2. Send them to the chosen AI provider (see global settings).
3. Receive an hour estimate with rationale.
4. Save the result in the local database (tracker_ai_estimates) — re-running is possible at any time.

Global module settings

Configured under Settings → Tracker:

• AI provider for estimates (trackerAiProviderId) — which provider from the AI Chat registry powers AI estimation.
• AI estimation prompt (trackerAiEstimatePrompt) — request template. Adapt it to your methodology (story points / hours / T-shirt sizing).

Polling interval (poll_minutes) is per-connection, not global.

Workflow

1. Connect a tracker

1. Open the Tracker module from the left sidebar.
2. In the sidebar click Add connection.
3. Pick the adapter type (Yandex Tracker / JIRA / GitHub / GitLab / Linear / Asana / Trello / Redmine).
4. Fill in the required fields (host, token, projects — depends on adapter).
5. Optionally configure status mapping — which source statuses map to internal ones.
6. Save. The first sync runs automatically.

2. Trigger a manual sync

Each connection in the sidebar has a Sync button. Without it the poll runs on the configured interval (typically 5–15 minutes, set per connection).

3. Work with tasks

• List — choose the default view to scan the queue.
• Board — drag cards between status columns.
• Filter — build your own conditions (e.g. "my tasks, in progress, critical priority").
• Task detail — click a row to open the right-side panel: description, comments, checklists, activity feed, links, AI estimation.

4. Run AI estimation

In the task detail switch to AI estimation → Run AI estimation. The result is saved and visible as an extra metric in the list.

5. Manage connections

• Edit — change parameters or status mapping.
• Delete — disconnect and remove the local cache for related tasks.

Next steps

• Configure AI providers — without them AI estimation does nothing.
• Use the Calendar to see tasks with deadlines on a timeline.
• For centralized code-project analytics, run Git Analytics alongside.

Help and feedback

Found a bug or want to request a new adapter? Contact us via the feedback form.