Priority Scoring for Jira — Documentation

In the second screenshot each field type appears twice; on your site you see each one once.

If your project uses separate screens for create, edit and view, add the fields to all of them, at least to the view screen, so scores are visible whenever you open an issue.

Scores are rounded to one decimal. Where Jira shows the field as plain text, for example in issue lists, the score has a colored dot with fixed breakpoints (RICE 3/10, WSJF 2/5, ICE 20/50) that does not follow your thresholds.

Board Health is kept for up to 1 hour. Scores set with the sliders or through Rovo, and changes to the Health Criteria, update it right away. Refresh recalculates on demand. Scores changed outside the app, for example through a script or the Jira REST API, show up after the hour or after Refresh.

If Jira refuses a search for a moment, the ranking and Board Health say what happened, with the HTTP status, instead of showing an empty project. If the app cannot check your issue permissions for a moment, it shows no issue list and says why (with the HTTP status). Click Refresh after a moment. For a project you cannot open, ask a Jira administrator for Browse Projects or choose another project.

SCRUM Defaults turns on Stale, Missing Description, Missing Story Points, Unscored and Overdue. KANBAN Defaults turns on Stale, Missing Description, Unscored, Missing Assignee, Missing Priority and Missing Labels. When Stale is on, a slider below it (7–180 days in steps of 7) sets the threshold for this project; it overrides the site-wide value from Settings.

Every enabled criterion counts equally: with 4 criteria each can subtract up to 25 points, with 5 up to 20, and so on, so the score always stays between 0 and 100. Board Health looks at up to 100 open issues, those not updated for the longest time first. A project with fewer open issues is analysed completely.

Framework names work in any case (WSJF, wsjf). A score is saved only when every dimension of that framework is given. An unknown framework, missing dimensions or values of another framework are rejected with a message, and nothing is written. If Jira refuses a search, for example because of a mistyped project key, the agent says so instead of returning an empty result.

Your Jira permissions apply: the agent only reads projects and issues you can browse and applies a score only to issues you may edit. Other issues stay unchanged and are named in the answer. The apply mode is checked again on every write: in Suggest only nothing is written and bulk scoring runs only in Auto, even from an older chat.

Rovo may split a large run into batches, and Atlassian shows its security dialog for every batch of write actions. This is normal; confirm each one.

An issue's entries are deleted 90 days after its score was last saved. Turning the trail off stops new entries; it does not delete old ones.

They are the denominator. A weight of ×2 there would halve the score, the opposite of what you expect. To de-emphasise effort, raise the other weights instead. In ICE all three dimensions can be weighted.

Low 0–20 / 0–5 / 0–100 and High 1–50 / 1–15 / 1–500 for RICE / WSJF / ICE.

• Columns: Project key and name, Health with grade, Issues analysed, Stale, Unscored and Last Checked (minutes ago).
• Search by project key or name and filter by Grade (All, Poor, Fair, Good+). A banner counts the projects rated Poor (below 40).
• Tick up to 5 projects to see their average score and totals. Analyse with Rovo then shows a message to paste into a Rovo chat.

Columns: Time, Issue (a link), Framework, Score and Mode (confirm or auto); 25 rows per page by default (10, 25, 50 or 100). For one issue's history, use the audit trail.

• Only open (not Done) issues are imported.
• A run covers the first 500 open issues of the project, oldest created first. If there are more, the dialog says how many are not covered and the button reads Import first 500 of N issues — import now. Need more? Contact support (pre-filled email).
• A run takes a few seconds, up to about 20.
• Import writes only to Priority Scoring's own fields. An existing Priority Scoring value of the imported framework is replaced, including scores set by hand. Other apps' fields and Jira's built-in fields are never changed.
• Import is only available on the Import tab, not through the Rovo agent, so the preview is never skipped.

The scan is a name-based guess. A field is offered if it is numeric, does not belong to Priority Scoring, and its name contains the whole word RICE, WSJF (or “weighted shortest”) or ICE. It can miss fields and offer unrelated ones (“ICE Cream Count” would match), which is why the preview shows the field name first.

• Detected: “RICE Score”, “RICE Priority”, “WSJF”, “WSJF Score”, “ICE Score”, and dimension fields such as “RICE Reach” or “RICE Effort”.
• Not detected: “Priority Score”, “Scoring Value”, “Custom Prioritization” (no framework word).

Score only vs. with dimensions: if the other app stored only a total, only the total is imported and Explain Priority Score has nothing to explain until you re-score the issue. If it stored separate dimension fields, the app takes those values and calculates the score with its own formula and your weights, so the result can differ from the old app's number.

The field is almost always missing from the project's screen. Go to Jira Settings → Work items → Screens, open the screen of your project and add the Priority Scoring field; in a team-managed project add it under Project settings → Issue types. See Add the fields to screens.

The Priority Scoring field is probably not on the project's screen. Rovo saves the values in the app, but Jira shows no value in the field until the field is on the screen. Fix the screen setup, then click Refresh on the app page.

The agent checks the framework and the dimension values before it writes anything. Name one of RICE, WSJF or ICE and give every dimension of that framework — RICE: Reach, Impact, Confidence, Effort; WSJF: Business Value, Time Criticality, Risk Reduction, Job Size; ICE: Impact, Confidence, Ease. Until the values are complete, nothing is written and the existing score stays as it is. In a bulk run, only the affected issues are skipped; the others are scored as usual. The agent also leaves an issue unchanged when you don't have permission to edit it in Jira, and names it in its answer.

You don't have permission to browse that project in Jira, or the project no longer exists. Priority Scoring checks this before it shows issues or scores, because it reads them with its own app permissions (see Only projects you can open). Ask a Jira administrator for the Browse Projects permission on the project, or edit the gadget and choose a project you can open.

Only Jira administrators can change Priority Scoring settings, because they apply to everyone on the site. Everyone else sees the current values read-only. Ask a Jira administrator to change a setting; the Health Criteria of a project can also be changed by that project's administrators.

Explain Priority Score uses the dimension values the app stored when the score was set with the sliders or by Rovo. A score set from outside (REST API, script) or imported as a total has no dimension values. Re-score the issue with the sliders or Rovo to enable explanations.

The agent reads the apply mode when a chat starts. Open a new Rovo chat after changing the mode. Every write is still checked against the current mode: in Suggest only, nothing is written, even from an older chat.

No. In Auto mode the agent skips its own question in the chat, but Atlassian always shows a security dialog before a write action. The app cannot turn it off. Confirm it.

Refresh recalculates right away. Board Health looks at up to 100 open issues, those not updated for the longest time first, so in a larger project a recently edited issue may not be in the sample. Also check that the score fields are on the project's screen and that the criteria you expect are enabled in Health Criteria.

The agent checks the 50 oldest open issues of the project for missing scores, because Jira cannot search these score fields. For a complete list use the Priority Ranking on the app page: it lists all open issues, and unscored ones show Not scored.

The app recognises the two most common story points fields of Jira Cloud without any setup. If your points are still not detected, contact support with your Jira site URL.

No. Existing scores are not recalculated. They update when a score is next saved on each issue, with a slider or by Rovo. After a big weight change, ask Rovo in Auto mode to re-score the issues in bulk.

RICE (Reach, Impact, Confidence, Effort) is popular for product and roadmap planning. WSJF (Weighted Shortest Job First) comes from SAFe and ranks by cost of delay per job size. ICE (Impact, Confidence, Ease) is a lightweight framework for quick scoring. Priority Scoring has a field for each of them.

The scores are values of the Priority Scoring fields in your Jira site. The app's own data (dimension values, settings, audit trail) is stored in Atlassian Forge storage for your site. The app runs on Atlassian's infrastructure and sends no data to external servers. Details in the Privacy Policy.

Go to Jira Settings → Apps → Manage apps, find Priority Scoring and click Uninstall. Atlassian keeps the app's stored data (scores, settings, audit trail) for 28 days after the uninstall and then deletes it, see the Privacy Policy. The custom fields stay in Jira until you delete them under Settings → Work items → Fields.