Playwright test automation with the uipath-test skill (Preview)

Bring an existing Playwright suite into UiPath Test Manager with the uipath-test skill, with no code changes required.

备注：

此功能处于预览阶段。预览版功能不支持用于生产，并且功能可能会受到限制。

The uipath-test skill lets a coding agent (Claude Code, Cursor, GitHub Copilot, Gemini CLI, or Codex) take an existing Playwright test suite, with no code changes, and bring it into UiPath Test Manager.

Ask your agent in plain language, and it creates the Test Manager project, packages the suite, uploads it to Orchestrator, auto-creates linked test cases, builds a test set, runs it, and reports pass, fail, and skip results.

When tests fail, you can ask the agent to fix them, repackage, republish, and re-run, entirely from the terminal. Without this skill, bringing an existing Playwright suite into Test Manager would otherwise require rewriting the tests as UiPath automations — this flow packages your Playwright code unchanged.

先决条件

• UiPath CLI (uip) installed and authenticated (uip login). CLI version 1.200.0 or later is required for the --playwright-projects flag used in this flow.
• UiPath skills installed for your coding agent (uip skills install --agent <agent>), which installs the uipath-test skill bundle.
• Playwright installed locally in the project you want to bring into Test Manager, installed with npm (a package-lock.json must be present). The skill packages your existing suite as-is and does not install Playwright for you.

Bring your suite into Test Manager

1. Open your coding agent in your Playwright project's root folder.

2. Ask it to bring the suite into Test Manager:

   "I have written some Playwright scripts to test my website [Product Name]. Can you help me run these tests using UiPath's Test Manager service?"

3. Answer the agent's setup questions on the first run: which Test Manager project the tests should land in (or whether to create a new one), and which Playwright scope (for example, regression, smoke, or checkout-critical) to execute first.

4. Review the pass, fail, and skip summary the agent reports, including which failures look like real assertion mismatches versus environment or infrastructure issues.

结果:

Behind the scenes, the agent creates the Test Manager project and its default Orchestrator execution folder, packages the Playwright project into a .nupkg, and uploads it to the Orchestrator package feed.

Test Manager auto-creates and auto-links one test case per discovered test. The agent then creates a test set scoped to the selected Playwright project, runs it, and polls until the execution finishes.

Fix a failing test and re-run

Ask your agent to fix and re-run:

"Fix the failing tests and re-run the smoke test set."

The agent fixes the failing test code, then repacks with a bumped package version, re-uploads, and re-runs the same test set. The Playwright project and scope selection persists on the test set, so you don't need to re-specify it.

警告：

Don't ask the agent to use uip tm executions retry to re-run after a fix. That command re-runs the previous execution's failures against the old package version, the opposite of this loop. Instead, the agent bumps --package-version, repacks and re-uploads, then re-runs the same test set.

命令参考

The agent selects these commands automatically from your prompts — you don't need to type them yourself. They're useful for troubleshooting, or for scripting the same flow in CI.

步骤	命令
Create the Test Manager project	uip tm project create --name "SwagLabs" --project-key SWAG --description "..." --output json
Set the default execution folder	uip tm project set-default-folder --project-key SWAG --folder-key <key> --output json
Package the Playwright project	uip tm pack --project-path . --type playwright --project-key SWAG --name SwagLabsPlaywright --package-version 1.0.0 -o .packages --output json
Upload to Orchestrator	uip or packages upload ".packages/SwagLabsPlaywright.1.0.0.nupkg" --output json
Poll for ingested test cases	uip tm testcases list --project-key SWAG --output json
Run the test set, scoped to a Playwright project	uip tm testsets run --test-set-key SWAG:37 --playwright-projects regression --output json
Poll execution status and results	uip tm report get --execution-id <id> --project-key SWAG --output json
Pull failure details for triage	uip tm executions testcaselogs list --execution-id <id> --project-key SWAG --only-failed --output json

备注：

uip tm testcases list takes no --limit flag.

另请参阅

• Tests — the full list of what your coding agent can do with Test Manager.
• uip tm pack — full reference for the packaging command used in this flow.
• uip tm testsets — full reference for testsets run and the --playwright-projects flag.
• uip or packages — full reference for publishing packages to Orchestrator.