Generate from a single requirement and its selected context.
AI test generation for Jira
AI test generation in Jira, with people in control
Generate test drafts directly from a Jira requirement. Requirement analysis is optional, every draft is editable, and nothing becomes a Jira Test Case until a person accepts it.
How it works
Draft tests without surrendering the decision
Nexus separates generation from acceptance. The model proposes cases and steps; your reviewer can edit, accept or dismiss each draft before Jira records are created.
Use step-based, data-driven and technique-assisted draft generation.
Keep optional ambiguity and gap analysis separate from test creation.
Review suggested steps, expected results and reusable data before acceptance.
Implementation guide
Bring your own AI account
A Jira administrator connects a supported Anthropic, OpenAI, Google Gemini, Moonshot Kimi or xAI Grok account for Nexus Pro. The provider key stays in Forge secret storage and model usage is billed under your provider agreement.
Site and project switches can disable AI.
Token history and warning thresholds support operational review.
Sensitive and masked dataset rows are excluded from AI examples.
AI output is never a release approval.
Questions buyers ask
Frequently asked questions
Must a requirement be analysed before generating tests?
No. Analysis is optional and is not a gate for test generation.
Does Nexus create Jira issues immediately?
No. Generation produces independent drafts. A person must explicitly accept a draft before Nexus creates the Jira Test Case.
Who pays for model usage?
Your organisation connects and pays its chosen AI provider separately from the Nexus Marketplace subscription.
Test the workflow with a real Jira project
Start with one requirement, one cycle and one release decision.