Compare

Keep the issue, backlog and report; skip the workflow engine.

Nothing Left is for teams that want issues, sprints, time logs and reports but do not need Jira’s extensive administration model: transition rules, permission schemes and field configurations to maintain. Five fixed roles, one search language and an MCP server that people and agents share.

Enough structure for the sprint. Less administration around it.

Nothing Left and Jira, row by row.

Both track issues, run sprints, log time and answer agents. Where Jira goes further, the note says how; where Nothing Left has nothing, the cell says No. Each cell links to the page it was checked against.

Jira column last verified against its own pages on . Product names are trademarks of their owners.

Nothing Left and Jira, row by row. Nothing Left compared with Jira, by capability.
CapabilityNothing LeftJira
Planning
Issue structureYesBug, Feature, Task and Epic by default, up to three levels deep, with links, tags and attachments. Project Admins rename or add types.Issues manualYesEpics, stories, tasks, bugs and subtasks; Jira admins create custom work types. Levels above epic come with Plans on Premium and Enterprise.Jira work items guide
Boards and sprintsYesBacklog, one running sprint per project, story points, boards by status, priority or type, swimlanes and WIP warnings.Sprints and boards manualYesScrum and kanban boards, sprint planning from the backlog, estimation and customisable board workflows.Jira agile page
Cross-team planningNoSprints and boards belong to one project. Reports and searches can span projects, but there is no multi-team plan, capacity view or dependency timeline.Reports manualYesPlans, on Premium and Enterprise, combine work from several teams and spaces with dependencies, capacity and what-if scenarios.Plans in Jira Premium
Custom fieldsYesPer project, from select and user to date and duration, each searchable and in the CSV export. Project Admins add them.Custom fields manualYesJira admins create fields of many types, then place them through field configurations and contexts.Create a custom field
PermissionsLimitedFive fixed roles from Observer to Org Owner. No custom roles, permission schemes or per-comment visibility.Permission matrixYesPermission schemes grant capabilities through space roles and can be shared across spaces. Not available on Free sites.Manage space permissions
Delivery
Query languageYesOne readable language, such as assignee: me #Unresolved, in the search bar, saved views, board filters, reports and agent tools.Search manualYesJQL builds structured searches from fields, operators and values in advanced search.Advanced search with JQL
ReportsYesNine reports, including burndown, time, Gantt and timesheet, each with a table under the chart and CSV export. No velocity or cumulative flow chart.Reports manualYesBurndown, velocity, cumulative flow and sprint reports for agile teams.Jira agile page
Time logsYesLog time on the issue against its estimate, with work types; subtasks roll up into the parent, and timesheets read the same entries.Time tracking manualYesOriginal estimate, logged time and time remaining on the work item, in weeks, days, hours and minutes.Log time in Jira Cloud
AutomationLimitedStatus rules fired by git events, resolved subtasks and duplicate links, such as Close on merge, from a fixed list of triggers; not a general rules engine. No transition rules: any status can follow any status.Automations manualYesA no-code rule builder of triggers, conditions and actions, with branches; usage limits vary by plan.Jira automation overview
IntegrationsLimitedGitHub and GitLab: branches, commits and pull requests linked to issues, read-only. No chat or outgoing webhook integrations yet; MCP, the REST API and CSV export are the other ways in and out.Development manualYesConnections to tools such as GitHub and Slack, plus apps from the Atlassian Marketplace.Atlassian Marketplace
Agents
MCP and agentsYesAn MCP server on the same public API as the app, over HTTP or stdio. A scoped personal token acts as you, within your role.MCP manualYesThe Atlassian Rovo MCP Server lets external AI clients search, create and update work items, with OAuth 2.1 or an API token.Rovo MCP Server docs
Hosting and pricing
HostingYesAGPL-3.0 code, the same build as the hosted service, with seat and storage limits as settings. Guides for Docker Compose and AWS.Self-hosting guideLimitedCloud, hosted by Atlassian. Self-managed Data Center closed to new customers on 30 Mar 2026 and reaches end of life on 28 Mar 2029.Data Center end of life
PlanYesNo subscription for any team size during beta; the first 20 members stay free when billing arrives. 5 GB of storage, or your own bucket.Limits and billingYesA Free plan for up to 10 users, then Standard, Premium and Enterprise. Current prices are on Atlassian’s pricing page.Jira plans
Yes
Built in and documented.
Limited
Present, with a narrower scope than the row suggests; the note says how.
Planned
Not available today; on the public roadmap or announced.
No
Not available.

From a new project to a burndown, in five screens.

What setting up looks like when there is no scheme to choose first. Acme’s team creates ACME, plans Sprint 08, finds its open bugs, moves one to Fixed and reads the chart. Pick a step to see that screen.

1 / 5An Org Admin creates the project

Acme/New project
Project nameAcme
Project keyACMEUsed in issue keys, such as ACME-1. It can change until the first issue exists.

Ready the moment it exists

Statuses
Submitted, Open, In Progress, Reopened, Can’t Reproduce, Duplicate, Fixed, Won’t fix, Obsolete, Closed
Priorities
Show-stopper, Critical, Major, Normal, Minor
Types
Bug, Feature, Task, Epic
Board
To do, In progress and Done, with a row per epic
Time tracking
On, with Development as the default work type
Roles
Observer, Contributor, Project Admin, Org Admin, Org Owner

No workflow, screen or field configuration to pick first. A Project Admin can rename, add or archive values later; any status may follow any status.

2 / 5Casey Kim plans Sprint 08 · 3 Oct 2026, 10:20am

ACME/Backlog

Sprint 08PlannedMake invitations clearer · 9 issues, 31 points · 4 Oct – 17 Oct 2026

  • ACME-12Invite link blank after it expires3 points
  • ACME-14Invite email subject names the org2 points
  • ACME-15Resend invite from the members list1 point
  • ACME-11Swimlane header overlaps count5 points
  • ACME-10Invite landing shows the inviter3 points
  • ACME-6Project-scoped invite links8 points
  • ACME-5Accept invite while signed in5 points
  • ACME-4Invite expiry setting per org3 points
  • ACME-2Password hint is cut off on small screens1 point

BacklogIssues in no sprint

  • ACME-3Make invitations clearerNot estimated
  • ACME-7Burndown ignores late additionsNot estimated
  • ACME-9Upload stalls at 100% on slow linksNot estimated
  • ACME-13Invite expiry date shown in UTCNot estimated
  • ACME-16Reminder email before an invite expires3 points
  • ACME-17Show who invited each member2 points
  • ACME-18Revoke a pending invitation5 points
  • ACME-19Invite form explains each role1 point

Putting an issue into a sprint is an ordinary issue edit, open to Contributors. Starting and finishing the sprint is for Project Admins.

3 / 5The sprint’s open bugs · 7 Oct 2026, 9:00am

ACME/Issues

#{Sprint 08} type: Bug #Unresolved sort by: updated desc6 issues

  • ACME-12Invite link blank after it expiresMajorAlex Lee20mIn Progress
  • ACME-5Accept invite while signed inMajorAlex Lee-In Progress
  • ACME-13Invite expiry date shown in UTCMinorAlex Lee-Open
  • ACME-9Upload stalls at 100% on slow linksMajorUnassigned-Open
  • ACME-11Swimlane header overlaps countNormalCasey Kim1h 30mIn Progress
  • ACME-2Password hint is cut off on small screensNormalJordan-Open

The same text filters the board and scopes a report, and an agent passes it to search_issuesunchanged. Save it as a view and it becomes a tab with its count.

4 / 5Alex Lee moves ACME-12 · 8 Oct 2026, 4:40pm

ACME/Board/Sprint 08

Make invitations clearer

Ends 17 Oct 2026

Open4

2ACME-14
Invite email subject names the orgCasey Kim
1ACME-15
Resend invite from the members list
2ACME-9
Upload stalls at 100% on slow links
1ACME-2
Password hint is cut off on small screensJordan

In Progress3

5ACME-11
Swimlane header overlaps countCasey Kim
3ACME-10
Invite landing shows the inviterAlex Lee
8ACME-6
Project-scoped invite linksJordan

Fixedresolved4

3ACME-12
Invite link blank after it expiresAlex Lee
5ACME-5
Accept invite while signed inAlex Lee
3ACME-4
Invite expiry setting per orgSam Ortiz
1ACME-13
Invite expiry date shown in UTCAlex Lee

Acme’s board uses its statuses as columns: Open with Reopened, In Progress, and Fixed. Fixed is a done status, so the move resolves the issue and its reporter is told. No transition screen, no validator.

5 / 5Burndown · Sprint 08 · 13 Oct 2026, 9:20am

ACME/Reports/Burndown

21 of 34 points resolved3 points added after the start4 days left

  • Remaining
  • Ideal
  • Scope
Daily figures
Sprint 08daily burndown figures, in story points
DayRemainingIdealScopeAddedDone
4 Oct31313100
5 Oct3427.934+30
6 Oct3124.834+33
7 Oct2521.734+39
8 Oct2218.634+312
9 Oct1415.534+320
10 Oct1415.534+320
11 Oct1415.534+320
12 Oct1312.434+321
13 Oct139.334+321
14 Oct-6.2---
15 Oct-3.1---
16 Oct-0---
17 Oct-0---

One of nine reports. Each has the table under its chart and a CSV export, and an agent reads the same numbers through run_report. Burndown in the manual

Hand-drawn product views with fictional data.

Five roles, decided before you arrive.

There is no role editor and no permission scheme to design. Each person has an organisation role and a role in each project, and the same matrix answers the app, the API and every agent.

That is a trade, not a shortcut to every permission need. Jira’s permission schemes and workflow conditions go much further.

The full permission matrix

The five roles

Observer
Reads issues, boards, reports and articles. Changes nothing.
Contributor
Files and edits issues, comments, logs time and moves issues into sprints.
Project Admin
Also runs sprints, shapes boards, statuses and fields, and manages project members.
Org Admin
Project Admin on every project, plus people, invitations and organisation settings.
Org Owner
Everything an Org Admin can, plus transferring ownership and deleting the organisation.

Deliberately not built

  • No custom roles and no permission schemes. Each role means the same thing in every project.
  • No field- or comment-level visibility: everyone who can open an issue sees all of it.
  • No transition conditions. Who may move an issue is decided by role, not by workflow.

What to check before you move.

Moving issues is the easy part. What decides whether a team can leave Jira is the configuration around them, so start with these four.

  1. Workflow rules

    List the transitions, conditions, validators and post functions your team relies on. None of them carry over: here any status can follow any status, so a rule that blocks or triggers a move becomes a team habit or an agent request.

    Jira workflow configuration
  2. Add-ons and automation

    Write down what each Marketplace app and automation rule does for you. Automations here only set a status, on git events, resolved subtasks and duplicate links, and there is no app directory, so decide which jobs you can drop and which would keep you on Jira.

    Jira automation overview
  3. Custom field mappings

    Match every field you use to a field type here, per project. Options, people and dates map by name; anything with no equivalent goes into the description or is left behind.

    Custom fields manual
  4. Historical data

    Decide how much history you need. Issues created here start their own history on the day they arrive, so past transitions and long comment threads are best left in Jira or quoted in the description.

    Issue history manual

The bulk path: CSV

Export work items from a Jira search, then use the generic CSV import: map the columns to issue fields, run a dry run and read the errors row by row before anything is created.

Export search results in Jira

Not built: a native Jira importer

Nothing reads a Jira site directly, and workflows, automation rules and app data have no place to land. Plan to leave them behind; a rule that moves an issue on merge can be set up again here.

The guided path: an agent on both sides

Only when one agent can reach your Jira site and Nothing Left. It reads the unresolved work there and creates issues here through the documented tools, after you approve the mapping.

Find the unresolved work items in our Jira space WEB, propose how people, statuses, types and fields map to the ACME project, and wait for my approval before you create anything.

  1. Connect both. An agent that reaches Jira through a connector you already use, and Nothing Left through a personal access token.
  2. Review the mapping. People to @handles, statuses and types by name, fields by type. Nothing is created until you approve it.
  3. Create and check. The agent creates issues, tags and links as you, then lists what it made. Spot-check them before you tell the team.
Connecting an agent

Which one fits your team.

Jira is the deeper system: cross-team planning, automation, a large app ecosystem and fine-grained control. Nothing Left is the smaller one, on purpose. Pick by the work you have, not by the longer feature list.

Choose Jira if…

Choose Nothing Left if…

  • One or a few product teams want issues, a backlog, sprints, time logs and reports without an administrator to configure them.
  • Five fixed roles cover who may read, edit and plan, and any status may follow any status.
  • People and AI agents should share one query language and one public API, with no extra layer between them.
  • You want the code under AGPL-3.0 and a path to running the same build yourself.

Sprints and boardsMCP and AI agents

Compare what Jira costs you now.

Jira is selected, starting from its lowest paid plan. Replace that with what your team pays today and see what the beta takes off the bill. Nothing you type leaves this page.

The calculator is not available in this browser right now. Reloading the page may bring it back. The short version: during beta the subscription is zero, whatever you pay today.

Currently using

Check your plan on Jira's pricing page

Currency

You know

members

Digits only, such as 1250, 1,250.50 or 1.250,50.

Filled in from Jira's lowest paid plan: US list price, billed monthly, checked 11 Oct 2026. Higher plans cost more, and annual or volume pricing differs. Replace it with your invoice figure.

Current subscription spend
-
Nothing Left subscription during beta
$0
Illustrative subscription spend avoided during beta if you replace and cancel the current plan
-

Enter what you pay today to see the comparison.

Monthly subscription only, in the currency you chose, with no exchange conversion. Filled-in prices are US list prices for each tool's lowest paid plan, billed monthly, checked 11 Oct 2026; they apply in USD only. Migration effort, add-ons, storage and any future seat pricing are outside this comparison. Product names are trademarks of their owners.

Questions about moving from Jira.

Import, agents and what needs a person to check it. Answers checked on 8 Oct 2026.

Can we import our Jira issues?

Use the generic CSV import. Export your work items from a Jira search as CSV, map its columns to issue fields in Nothing Left, run a dry run, and read the per-row errors before anything is created. There is no native Jira importer, and none is built today. Export search results in Jira

What can an agent carry over from a connected Jira account?

If one agent can reach your Jira site and Nothing Left, it can read unresolved work items there and use Nothing Left’s tools to create the project, invite members, create tags and create issues with title, description, type, priority, status, assignees, story points, estimate, sprint, parent, links and custom field values.

It works with your access on both sides and creates everything as you. Treat it as a guided move you review, not a one-click import. Connecting an agent

What needs manual review?

Who is who across the two accounts, how statuses, types and fields map, duplicates, and attachments, which have to be uploaded again. Original reporters, creation dates, comments’ authors and the change history do not transfer as data; the agent can quote them in the description if you ask.

Workflow rules and app data have no equivalent and stay behind. Automation rules do not carry over; one that moves an issue when a pull request merges can be set up again here as an automation.

Do our Jira workflows come across?

No. Status names can match, but transitions, conditions, validators and post functions are not modelled: any status can follow any status. Plan the move around that rather than expecting the rules to be rebuilt. Statuses and resolution

Will five roles replace our permission schemes?

Only if your needs are simple. The roles decide who reads, edits, plans and administers, the same way in every project. If you rely on custom roles, field or comment visibility, or conditions on who may move work, those are deliberately not built. The permission matrix

Run one sprint here before you decide.

Join the beta, create a project and plan a sprint with the work you have now. Keep Jira running beside it until you know which one the team would rather open.