--- title: Show, don't prompt: how QA Wolf turns a recording into a... description: QA Wolf added a recorder to its testing agent that turns a clicked-through flow into a real Playwright test. You record in a live browser next to the agent's... canonical: https://daily.dev/posts/show-don-t-prompt-how-qa-wolf-turns-a-recording-into-a-real-test-kcwrakwbn twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Show, don't prompt: how QA Wolf turns a recording into a real test | daily.dev og:description: QA Wolf added a recorder to its testing agent that turns a clicked-through flow into a real Playwright test. You record in a live browser next to the agent's... og:url: https://daily.dev/posts/show-don-t-prompt-how-qa-wolf-turns-a-recording-into-a-real-test-kcwrakwbn og:image: https://api.daily.dev/og/posts/KcWrAKWbn.png og:image:alt: Show, don't prompt: how QA Wolf turns a recording into a real test og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Show, don't prompt: how QA Wolf turns a recording into a real test **[QA Wolf](https://daily.dev/sources/qawolf)** · 10 min read · 0 upvotes · 1 comments ## Summary QA Wolf added a recorder to its testing agent that turns a clicked-through flow into a real Playwright test. You record in a live browser next to the agent's chat, and it keeps a shared step list that both of you can edit. Noisy browser events get collapsed, for example 11 raw events during a sign-in become just three steps: type email, type password, click Sign in. Before writing code, the agent replays the steps from a clean browser state, retrying a step up to three times a second apart to tell a real failure from a slow page. Once replay passes, the steps are rewritten into the same page-object style (@qawolf/pom) as any other test the agent writes. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How does QA Wolf turn a recorded browser session into a Playwright test? A recorder captures every click and keystroke in a live browser, then two cleanup passes reduce noise: the recording script merges rapid key presses into single fills, and a set of capture rules strips corrections, double-fired clicks, and redirect hops. For example, 11 raw events during a login collapse into three steps. The agent then replays the steps from a clean browser state before writing final Playwright code in a page-object style. _Teams automating UI flows can track approaches like step-recording to Playwright code on daily.dev._ ### How does an AI test agent tell a genuinely broken step from a page that's just still loading during replay? Each non-check step gets up to three retry attempts spaced a second apart, since a page still loading can cause a good step to fail once. A click that opens a slow page but already moved the app forward counts as passed rather than being retried, preventing duplicate clicks. Only after all retries fail does the replay stop and report the specific missing or broken step. _Developers debugging flaky end-to-end tests can follow retry and replay strategies like this on daily.dev._ ### What code style does QA Wolf's agent use when generating tests from a recorded flow? Recorded steps are rewritten into the page-object style used by the '@qawolf/pom' package, the same style the agent uses for tests it writes on its own without recording. Each locator lives once inside a page object, and each action becomes a named method returning the page it leads to, reusing page objects already present in the suite instead of creating duplicates. _Developers standardizing test suite architecture can compare page-object patterns like this on daily.dev._ ## Community discussion Top comments from developers on daily.dev. **@demovexa** · 0 upvotes > Replaying from a clean browser state before any code gets written is the step that makes this trustworthy. A recording that only passes in the session where it was made isn't a test yet. The collapsed step list is useful beyond testing too. Three named steps in place of eleven raw events are already close to the captions you'd want for a how-to clip of the same flow, and since the agent keeps that list current, it could flag which walkthrough videos or docs screenshots show a flow whose steps just changed. That's usually something teams only find out about from a confused user. ## Similar posts on daily.dev - [Let your AI agent capture its changes with the Playwright CLI](https://daily.dev/posts/let-your-ai-agent-capture-its-changes-with-the-playwright-cli-vt9zboanx) · Tim Deschryver · 1 upvotes · 3 comments - [How to Make Your Coding Agent Check Its Work in the Browser](https://daily.dev/posts/how-to-make-your-coding-agent-check-its-work-in-the-browser-faxnmjz0d) · freeCodeCamp · 2 upvotes · 2 comments - [Why your coding agent shouldn't be your QA system](https://daily.dev/posts/why-your-coding-agent-shouldn-t-be-your-qa-system-p3wg9nmqm) · QA Wolf · 2 upvotes · 1 comments - [Fire Your QA Today: Automate QA testing with AI](https://daily.dev/posts/fire-your-qa-today-automate-qa-testing-with-ai-7jyqjppkp) · Product Hunt · 0 upvotes · 0 comments - [Record a product demo with Playwright](https://daily.dev/posts/record-a-product-demo-with-playwright-bfaidl1pm) · Flavio Copes · 0 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#testing](https://daily.dev/tags/testing) [View this post on daily.dev](https://daily.dev/posts/show-don-t-prompt-how-qa-wolf-turns-a-recording-into-a-real-test-kcwrakwbn) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Show, don't prompt: how QA Wolf turns a recording into a real test","url":"https://daily.dev/posts/show-don-t-prompt-how-qa-wolf-turns-a-recording-into-a-real-test-kcwrakwbn","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/show-don-t-prompt-how-qa-wolf-turns-a-recording-into-a-real-test-kcwrakwbn"},"datePublished":"2026-10-28T20:20:57.954Z","dateModified":"2026-10-28T21:35:45.796Z","description":"QA Wolf added a recorder to its testing agent that turns a clicked-through flow into a real Playwright test. You record in a live browser next to the agent's...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/6bd135dfe46ce0023b5e384a5268e1e1?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/6bd135dfe46ce0023b5e384a5268e1e1?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"QA Wolf","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"QA Wolf","logo":"https://media.daily.dev/image/upload/s--B9OvVIWd--/f_auto,q_auto/v1788684983/logos/qawolf?_a=BAMAMicg0","url":"https://daily.dev/sources/qawolf"},"commentCount":1,"discussionUrl":"https://daily.dev/posts/show-don-t-prompt-how-qa-wolf-turns-a-recording-into-a-real-test-kcwrakwbn","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":1}],"keywords":"ai-agents,testing","timeRequired":"PT10M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"QA Wolf","item":"https://daily.dev/sources/qawolf"},{"@type":"ListItem","position":3,"name":"Show, don't prompt: how QA Wolf turns a recording into a real test"}]} {"@context":"https://schema.org","@type":"WebPage","@id":"https://daily.dev/posts/show-don-t-prompt-how-qa-wolf-turns-a-recording-into-a-real-test-kcwrakwbn","comment":[{"@type":"Comment","text":"Replaying from a clean browser state before any code gets written is the step that makes this trustworthy. A recording that only passes in the session where it was made isn’t a test yet. The collapsed step list is useful beyond testing too. Three named steps in place of eleven raw events are already close to the captions you’d want for a how-to clip of the same flow, and since the agent keeps that list current, it could flag which walkthrough videos or docs screenshots show a flow whose steps just changed. That’s usually something teams only find out about from a confused user.","datePublished":"2026-10-30T03:04:07.110Z","url":"https://daily.dev/posts/KcWrAKWbn#c-uPnTUv0Gv","author":{"@type":"Person","name":"DemoVexa","url":"https://daily.dev/demovexa","image":"https://media.daily.dev/image/upload/s--I0Ym8soj--/f_auto,pg_1/v1792077531/avatars/avatar_F6wD6jMSPjwEMcztjwnyE?_a=BAMAMicg0"}}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/show-don-t-prompt-how-qa-wolf-turns-a-recording-into-a-real-test-kcwrakwbn#faq","mainEntity":[{"@type":"Question","name":"How does QA Wolf turn a recorded browser session into a Playwright test?","acceptedAnswer":{"@type":"Answer","text":"A recorder captures every click and keystroke in a live browser, then two cleanup passes reduce noise: the recording script merges rapid key presses into single fills, and a set of capture rules strips corrections, double-fired clicks, and redirect hops. For example, 11 raw events during a login collapse into three steps. The agent then replays the steps from a clean browser state before writing final Playwright code in a page-object style. Teams automating UI flows can track approaches like step-recording to Playwright code on daily.dev."}},{"@type":"Question","name":"How does an AI test agent tell a genuinely broken step from a page that's just still loading during replay?","acceptedAnswer":{"@type":"Answer","text":"Each non-check step gets up to three retry attempts spaced a second apart, since a page still loading can cause a good step to fail once. A click that opens a slow page but already moved the app forward counts as passed rather than being retried, preventing duplicate clicks. Only after all retries fail does the replay stop and report the specific missing or broken step. Developers debugging flaky end-to-end tests can follow retry and replay strategies like this on daily.dev."}},{"@type":"Question","name":"What code style does QA Wolf's agent use when generating tests from a recorded flow?","acceptedAnswer":{"@type":"Answer","text":"Recorded steps are rewritten into the page-object style used by the '@qawolf/pom' package, the same style the agent uses for tests it writes on its own without recording. Each locator lives once inside a page object, and each action becomes a named method returning the page it leads to, reusing page objects already present in the suite instead of creating duplicates. Developers standardizing test suite architecture can compare page-object patterns like this on daily.dev."}}]} ```