Free and open-source · MIT

AI-ready Xray workflows for manual QA and automation engineers.

jiraxray-cli turns test intent into the right Xray tests, suites, executions, evidence, and reports — with AI assisting inside a controlled command workflow.

Open the usage guide See the workflow
controlled workflow jiraxray-cli 
# Discover what the CLI can do.
$ jiraxray-cli commands --json
ok known commands, inputs, risks, and next steps

# Review the plan before anything changes in Xray.
$ jiraxray-cli suite run regression-sync --dry-run --json
review planned actions only

# Approved work updates Xray repeatably.
$ jiraxray-cli results import --format junit --path report.xml --json
ok result import queued

The problem it solves

A basic MCP server can expose tools to an AI assistant. jiraxray-cli goes further: the same controlled actions can be run by a tester, a release script, or an existing automation suite — so the tool boundary stays clear no matter who starts the work.

Where it sits

Xray sits on Jira. jiraxray-cli sits beside both.

The CLI complements the Jira/Xray stack — it doesn't replace either. It gives testers, automation, and AI tools a controlled way to work alongside them.

Atlassian Jira Cloud

Issues, projects, plans, executions, and team workflow live here.

Xray Cloud

Adds tests, runs, evidence, results, sets, and plans on top of Jira.

jiraxray-cli

A controlled command layer for humans, automation suites, and AI assistants.

service boundary 
# Jira and Xray remain the source of truth.
$ jiraxray-cli issues summary --issue QA-123 --json
$ jiraxray-cli tests search --jql "project = QA" --json

# Read first. Write only by explicit command.
review no Xray changes happen during read.
Skip workflow animation

The workflow it supports

Manual / functional QA

Manual tests still belong in Xray.

For scenarios where a tester owns the wording, expected result, data, and evidence. AI can help shape the test, but the tester keeps the final say.

01
Test ideaScenario, data, and expected result.
02
Draft stepsTurn intent into Xray-friendly wording.
03
Dry-runPreview the exact Xray change.
04
Tester approvesFix wording or mapping before write.
05
Create in XrayTests, steps, sets, or plans update.
command workflow path 
# Same shape, different starting point.
$ jiraxray-cli commands --json            # discover
$ jiraxray-cli suite run regression-sync --dry-run
review planned steps only — no Xray changes

# After review:
$ jiraxray-cli suite run regression-sync --json
ok tests, executions, evidence, results aligned

Scroll to follow each path, or click a tab to jump. Manual-only tests, AI-prepared work, and automation imports all use the same controlled boundary.

AI without lock-in

More versatile than a basic MCP server.

MCP is useful when an AI assistant needs tools. jiraxray-cli keeps that structured tool behaviour, but it isn't trapped inside one AI host — humans and CI use the same commands.

AI-assisted

01
Discover

Run commands --json to learn the allowed actions.

02
Plan

Pick a safe command or run a dry-run first.

03
Explain

Show the planned action to the tester in plain English.

Human-approved

04
Review

Check the dry-run output and any planned writes.

05
Run

Approve an explicit write command — never automatic.

06
Reuse

Capture the flow as a versioned suite for next time.

Open source · MIT

Free to run, inspect, and adapt — open the guide and run the first dry-run.

jiraxray-cli is plain enough for test engineers to trust, structured enough for AI tools to use, and open enough for teams to adapt to their own Xray process.

Read the usage guide View on GitHub
  • AI tools can use it through terminal commands.
  • Humans can run the same commands manually.
  • CI jobs can run them automatically.
  • Workflow files can be reviewed and versioned.