Skip to content

Project Management Plan

Anushka Wijesundara edited this page Oct 14, 2026 · 4 revisions

Project management plan

Revised 2026-10-14. Goal: every project management capability of Jira Software, built into ServiceOps as a module of its own and connected to incidents, problems and changes. The owner narrowed the scope on 2026-10-14 to Jira only (no ClickUp target), and to its project management section:

  • Out of scope: Jira Service Management (ServiceOps is the ITSM platform), Confluence, Jira Product Discovery, Compass and the Atlassian Marketplace.
  • Kept from phase 1: features beyond Jira, such as multiple assignees and the project overview, stay as they are.

Backlog entries: B-443 (phase 1, done) and B-449 to B-459 (phases 2 to 12). Status is tracked in the backlog and the implementation plan.

Principles

  1. A module of its own, connected to ITSM. Work items link both ways to incidents, problems and changes.
  2. Reuse the platform. Tenancy, roles, audit, notifications, attachments and storage modes, live updates, search, the REST API with OpenAPI, AI, translations and themes all apply.
  3. Off until enabled. PROJECTS_ENABLED turns the module on.
  4. Each phase is a release. Every phase gets the full QA:
    • the full suite;
    • browser and accessibility checks;
    • stability fuzz on SQLite and PostgreSQL;
    • the cross-tenant sweep;
    • timing at 500 users and at least 10,000 work items per project.
  5. Jira's model, ServiceOps' rules. Jira concepts are kept (schemes, JQL, boards, sprints, versions, components) so Jira users feel at home and Jira data can be imported. Authorization, audit and tenancy follow ServiceOps.

Phases

Phase Backlog Content Status
1 B-443 Projects, work items, comments, history, attachments, links, overview, board, list, search, notifications, REST API Done (branch)
2 B-449 Work item completeness: resolution, components, versions, votes, flag, environment, time tracking and work logs, clone, move, convert, comment editing and visibility, more link types, rich text, Your work, starred and recent, keyboard shortcuts, bulk change Done (branch)
3 B-450 JQL, the issue navigator, saved and shared filters, favourite filters, filter subscriptions, CSV export Next
4 B-451 Scrum and Kanban boards (several per project, filter-based), backlog, sprints (including parallel sprints), estimation Planned
5 B-452 Reports: burndown, burnup, sprint, velocity, cumulative flow, control chart, epic, version, release burndown, created vs resolved and the issue reports Planned
6 B-453 Timeline: epics and child items, dependencies, drag to schedule, releases and sprints Planned
7 B-454 Workflow engine (transitions, conditions, validators, post functions), custom issue types, custom fields, screens, field configurations and the schemes that bind them Planned
8 B-455 Permission schemes, custom project roles, issue security levels, notification schemes Planned
9 B-456 Dashboards and gadgets Planned
10 B-457 Automation rules Planned
11 B-458 Development integration and smart commits, Jira import, CSV import, webhooks, REST API parity, project templates, trash, AI Planned
12 B-459 Plans (Jira's Advanced Roadmaps): cross-project plans, hierarchy above epic, capacity, scenarios Planned

The order follows dependencies:

  • boards are built on filters and JQL;
  • reports read sprints and versions;
  • automation needs the workflow engine.

Jira inventory

Status: Done (built), Pn (phase that builds it), Adapted (provided differently, with the reason).

Projects and administration

Jira capability Status
Project key, name, description, lead, avatar or colour Done
Private and open projects Done
Archive and restore a project Done
Project category P2
Default assignee (unassigned, project lead, component lead) P2
Components with component lead and default assignee P2
Versions: create, edit, start and release dates, release (moving unresolved items), archive, merge, delete P2
Project roles: administrators, members, viewers Done
Custom project roles P8
Project templates: Scrum, Kanban, bug tracking P11 (Scrum and Kanban defaults come with P4)
Trash: deleted projects kept 60 days before removal P11
Project features on/off (backlog, sprints, estimation, reports, releases, components, code) P4
Team-managed (simplified) and company-managed (scheme-based) projects Adapted: projects start simple, and schemes become available per project in P7 and P8
Global settings: issue types, priorities, resolutions, statuses, link types, time tracking P2 (resolutions, link types, time tracking), P7 (the rest)

Work items (issues)

Jira capability Status
Epic, story, task, bug, subtask Done
Custom issue types, issue type schemes, icons P7
Summary, description, priority, labels, due date, start date, parent Done
Assignee Done (several assignees allowed)
Reporter Done (changeable in P2)
Resolution (set on done, cleared on reopen) P2
Components, fix versions, affects versions P2
Environment P2
Story points and original estimate Done
Remaining estimate and time spent P2
Watchers Done
Votes P2
Flag (impediment) with comment P2
Rich text: headings, lists, tables, code, links, mentions, emoji, pasted images P2 (Markdown with a toolbar and preview)
Create another, inline edit Done
Assign to me P2
Clone (with or without subtasks, attachments and links) P2
Move to another project or type P2
Convert between subtask and standard type P2
Delete Done
Link types: blocks, relates, duplicates Done
Link types: clones, causes; custom link types P2
Links to ITSM incidents, problems, changes Done (beyond Jira)
Log work: time spent, date, comment, remaining estimate handling P2
Activity tabs: all, comments, history, work log P2 (comments and history done)
Edit and delete comments; comment visible to a project role only P2
Emoji reactions on comments P2
Share by notification; print view; export to Word or XML P2 (share and print); Adapted: CSV and JSON export instead of Word and XML (P3)
Attachments (drag and drop, preview) Done (drag and drop and image preview in P2)
Development panel (branches, commits, pull requests, builds, deployments) P11

Search and filters

Jira capability Status
Quick search by key and text Done
Basic search (field pickers) Done on the list; becomes the navigator's basic mode in P3
JQL fields and custom fields P3 (custom fields after P7)
JQL operators: =, !=, >, >=, <, <=, ~, !~, IN, NOT IN, IS EMPTY, IS NOT EMPTY P3
JQL history operators: WAS, WAS IN, WAS NOT, CHANGED with FROM, TO, BY, DURING, BEFORE, AFTER P3
JQL functions P3
JQL AND/OR/NOT, parentheses, ORDER BY P3
JQL autocomplete and syntax errors with positions P3
Saved filters with owner, shared with a project, role or everyone; favourites P3
Filter subscriptions (scheduled email) P3
Issue navigator: list and detail views, choosing columns P3
Export to CSV (all fields or current fields) P3
Bulk change: edit, transition, move, delete, watch, stop watching P2 (edit, transition, delete, watch), P3 (from any filter result)

JQL functions in P3:

  • people: currentUser(), membersOf();
  • sprints: openSprints(), closedSprints(), futureSprints();
  • versions: releasedVersions(), unreleasedVersions(), latestReleasedVersion(), earliestUnreleasedVersion();
  • time: now(), startOfDay(), endOfDay(), startOfWeek(), endOfWeek(), startOfMonth(), endOfMonth(), startOfYear(), endOfYear();
  • links and history: linkedIssues(), issueHistory(), watchedIssues(), votedIssues(), updatedBy(), parentEpic(), issuesWithRemoteLinks().

Boards, backlog and sprints

Jira capability Status
One board per project Done
Several boards per project; boards across projects from a filter P4
Scrum and Kanban board types P4
Columns mapped to one or more statuses P4 (today one column per status)
Column constraints: minimum and maximum, by item count or story points P4 (maximum by count done)
Swimlanes: none, assignee, epic Done
Swimlanes: stories (parent), projects, JQL queries P4
Quick filters (JQL) P4 (fixed filters done)
Card layout (extra fields) and card colours (by type, priority, assignee or query) P4
Days in column on cards P4
Detail panel on the board P4
Kanban backlog; released work hidden from the board P4
Estimation statistic (story points, original estimate, item count) and time tracking statistic P4
Working days and non-working dates for charts P4
Backlog: drag ranking, epics and versions panels, sums per sprint P4
Sprints: create, edit, start (name, duration, dates, goal), complete (move unfinished work to backlog or another sprint) P4
Parallel sprints P4
Sprint scope change tracking P4

Reports

Jira capability Status
Burndown, burnup, sprint report, velocity P5
Cumulative flow diagram P5
Control chart (cycle time) with rolling average P5
Epic report, epic burndown P5
Version report, release burndown P5
Created vs resolved, pie chart, recently created, resolution time, average age, time since P5
Single-level group by, user workload, time tracking report P5
Sprint health on the board P5

Timeline (roadmap)

Jira capability Status
Epics and child items as bars; collapse and expand P6
Dependencies drawn between bars; create by dragging P6
Drag and resize to change start and due dates P6
Progress bars, release and sprint markers P6
Week, month and quarter scales; filters; today marker P6

Workflows, fields and schemes

Jira capability Status
Statuses with categories, per project Done
Workflows with named transitions P7
Global transitions ("from any status") P7
Transition screens P7
Conditions: permission, project role, only assignee, only reporter, subtasks done P7
Validators: required fields, permission, date ordering P7
Post functions: assign, set or clear a field, set resolution, add comment, notify P7
Workflow per issue type through a workflow scheme P7
Workflow editor (diagram and text views) P7
Custom fields: text, paragraph, number, date, date-time, select, multi-select, radio, checkboxes, cascading select, user, multi-user, group, URL, version, labels, project P7
Field configurations (required, hidden, description) P7
Screens and screen schemes (create, edit, view) P7
Issue type and priority schemes P7

Permissions, security and notifications

Jira capability Status
Fixed roles: administer, edit, comment, view Done
Permission schemes: per-permission grants to roles, groups, users, reporter, assignee P8
Permissions covered: browse, create, edit, transition, assign, assignable, resolve, close, delete, link, move, schedule, set security, comments (add, edit own and all, delete own and all), attachments (create, delete own and all), work logs (add, edit and delete own and all), manage sprints, manage watchers, view voters and watchers, administer P8
Issue security levels and schemes P8
Notification schemes: events × recipients (reporter, assignee, watchers, role, group, user, component lead) P8
Personal notification preferences per event P2

Dashboards

Jira capability Status
Personal and shared dashboards, layouts, several per user P9
Gadgets: filter results, assigned to me, activity stream, pie chart, created vs resolved, two-dimensional filter statistics, issue statistics, heat map, calendar, sprint burndown, sprint health, average age, resolution time, watched issues, voted issues, road map, labels P9

Automation

Jira capability Status
Rules per project or across projects; rule audit log P10
Triggers: created, transitioned, field changed, commented, assigned, work logged, scheduled (with JQL), sprint started or completed, version released, manual, incoming webhook P10
Conditions: JQL, field compare, user, related items, if/else P10
Actions: transition, edit fields, assign, comment, create, clone, link, log work, notify, send email, webhook, set variable P10
Branches: subtasks, parent, epic, linked items, JQL P10
Smart values ({{issue.key}}, {{issue.summary}}, …) P10
Rule templates P10

Integration, import and API

Jira capability Status
REST API: projects, work items, comments Done
REST API: search (JQL), transitions, work logs, attachments, links, watchers, votes, components, versions, boards, sprints, filters, dashboards Matching phases (P2 to P9)
Webhooks for work item events P11
Development tools: link commits, branches and pull requests by key (GitHub, GitLab); smart commits (WEB-12 #comment … #time 2h #done) P11
Jira import (JSON or CSV export, keeping keys, links, comments, history, attachments, sprints and versions) P11
CSV import with field mapping P11
Email to work item P11
AI: summarise a work item and its comments, natural language to JQL, break an epic into stories P11
Mobile: work items in the iOS app P11

Plans (Advanced Roadmaps)

Jira capability Status
Plans across projects and boards P12
Hierarchy levels above epic (for example initiative) P12
Teams and capacity per sprint P12
Scenarios and auto-scheduling P12
Dependencies report P12

Your work and productivity

Jira capability Status
Your work: worked on, viewed, assigned to me, starred P2 (assigned to me done)
Starred projects, filters and boards; recent items P2 (projects and items), later phases (filters, boards)
Keyboard shortcuts P2
Global create button P2

Data model

Phase 1 tables:

  • project, project_member, project_status;
  • work_item with work_item_assignee, work_item_watcher, work_item_label;
  • work_item_comment, work_item_history, work_item_link;
  • file_attachment.work_item_id.

Phase 2 adds:

  • project_component;
  • project_version with work_item_fix_version and work_item_affects_version;
  • work_item_worklog, work_item_vote, work_item_component, work_item_star, work_item_view;
  • comment_reaction, link_type, resolution;
  • new columns on work_item: resolution, environment, flag, remaining and spent time;
  • new columns on project: category and default assignee.

Later phases add sprints, boards, filters, subscriptions, workflows, transitions, custom fields, screens, schemes, permission and notification schemes, dashboards, gadgets, automation rules and plans, each with tenant_id.

Acceptance for each phase

  • Each new behaviour has tests that fail without it, including negative authorization and other-tenant cases.
  • Browser QA at desktop and phone width, with axe WCAG 2.1 AA and no violations, in light and dark themes.
  • Stability fuzz reports 0 failing routes on SQLite and PostgreSQL.
  • The cross-tenant sweep is clean.
  • Every page stays under about 150 ms at 500 users and 10,000 work items per project on PostgreSQL.
  • The OpenAPI document is complete; a reversible migration is verified on PostgreSQL.
  • Documentation is updated.

Clone this wiki locally