Definition of Ready AI for Jira

Readiness scoring

Definition of Ready AI for Jira combines deterministic checks with AI-generated analysis so teams can see why an issue is ready, needs clarification, or should be blocked before implementation starts.

1. Statuses

  • Ready: the issue meets the configured score threshold and has no blocking critical findings.
  • Needs clarification: the issue has missing or ambiguous information, but the work is not clearly blocked.
  • Blocked: the issue has a low score or a blocking critical finding that should be resolved before implementation.

2. Scoring model

The readiness score starts at 100. Findings reduce the score based on severity:

  • Critical: minus 25
  • Warning: minus 12
  • Info: minus 4

The score is clamped between 0 and 100. Project admins can configure the minimum score required for an issue to be considered ready.

3. Common findings

  • Missing acceptance criteria
  • Unclear expected behavior
  • Missing scope boundary
  • Missing test expectations
  • Missing component or system context
  • Missing priority, design/spec link, dependency notes, rollout notes, or rollback notes
  • Risk-sensitive work without explicit review
  • Work that appears too large or broad

4. Risk-sensitive work

Issues involving authentication, authorization, payments, billing, data migrations, external APIs, sensitive data, rollout risk, or rollback risk may receive critical findings unless the issue includes enough review and mitigation detail.

5. AI and deterministic validation

The backend runs deterministic readiness checks and merges them with AI-generated findings. Teams should treat the output as review assistance, not as an autonomous approval decision.