Take a Jira story from requirement to review-ready QA coverage.
Use this guide to connect your sources, analyze feature evidence, generate test cases, and keep every QA decision traceable.
Quick start
Your first test plan in five steps
Start with one real Jira issue. VerifAI keeps generated work in draft until you choose to publish it.
Connect your source
Bring a Jira requirement into VerifAI
VerifAI reads the issue as product evidence. Jira text is never treated as an instruction to the agent.
If board synchronization is unavailable, load stories by issue key. Direct issue loading is enough to run analysis and test generation.
Grounded knowledge
Add design evidence from Google Drive or Microsoft 365
VerifAI uses delegated access to search only the documents your signed-in account can read. Imported text is versioned, chunked, and attached to the active QA project.
VerifAI requests read-only Drive access. Provider access tokens are used for the active session and are not written to VerifAI database tables or application telemetry.
Requirement analysis
Turn source evidence into QA decisions
VerifAI separates what is explicitly stated from what is missing, ambiguous, or risky.
How to run it
In Requirements, select the story and choose Analyze. Review the confidence score together with the evidence. A high confidence score can mean VerifAI is highly confident that the requirement is incomplete—not that the feature is ready.
Test design
Generate traceable test cases
Generated plans are drafts designed for human review. They remain grounded in the Jira issue and available supporting documents.
Review scenarios
Check the proposed positive, negative, boundary, integration, and regression coverage.
Generate cases
Select Generate test cases to produce steps, expected results, priority, and automation guidance.
Approve or publish
Select the useful cases. Create Jira issues only after the plan is ready for your team.
Publishing is deliberate: VerifAI does not create Jira issues merely because a plan was generated. Use Create Jira issuesonly after reviewing the selected cases.
Workspace reference
Know where each QA activity lives
Use the left navigation to move from sprint intake through release evidence.
Command Center
See sprint health, risks, active work, and release confidence.
Agent Runs
Review autonomous QA jobs, their steps, evidence, and outcomes.
Sprints
Load Jira issues, synchronize boards, and choose the active story.
Requirements
Review acceptance coverage, risks, questions, and testability gaps.
Knowledge
Manage design documents and supporting product evidence.
Test Design
Generate, inspect, and select traceable test cases.
Governance
Approve sensitive AI actions and preserve a decision trail.
Test Suites
Organize sanity, sprint regression, and full regression coverage.
Automation
Assess automation suitability and link automated coverage.
Executions
Track test runs, results, failures, and retest progress.
Defects
Triage failures and prepare evidence-based Jira defect drafts.
Release
Review readiness, risks, sign-offs, and release recommendations.
Traceability
Follow coverage from source requirement to test and result.
Insights
Inspect quality trends, coverage movement, and recurring risks.
Review and governance
Keep people in control of consequential actions
Treat VerifAI output as evidence-backed draft work. Review requirements, generated cases, defect drafts, and release advice before publishing.
Troubleshooting
Resolve common setup and workflow issues
Most problems can be isolated to connector access, missing source detail, or temporary AI availability.
Jira shows “Setup” or asks you to reconnect
Open Connectors, choose Connect Jira, select the correct Atlassian site, and approve the requested access. Reconnect if your authorization has expired.
A Jira issue cannot be loaded
Confirm the issue key and that your Jira account can view it. Use the full project key and number, for example SCRUM-2.
Board synchronization fails
You can continue by loading an issue directly from the Sprints page. Board synchronization also requires Jira Software access for that board.
AI features are unavailable or rate limited
Wait briefly and retry. If the message says AI is not configured or the credentials were rejected, ask your workspace administrator to check the server-side AI configuration.
The analysis says the story is not testable
Use the questions and testability gaps produced by VerifAI to refine the Jira description. Add explicit behavior and acceptance criteria, reload the issue, and analyze it again.
Ready to analyze a real story?
Return to the workspace, connect Jira, and start with one issue key.