Retracio
Join the waitlist
Menu

Checksum alternative: Retracio

Checksum is a cloud platform: an agent generates and heals tests in cloud sessions, reads your repository through a repo mirror, and delivers Playwright files to that repository as pull requests (checked 2026-09-08). Those files import Checksum fixtures, replaceable with standard Playwright imports (checked 2026-09-08). Retracio is a desktop app. You walk through a scenario by hand, mark what should be true on each screen, and it writes a plain Playwright spec into your local checkout, reusing the page objects, fixtures, and locator style already there. The model is called through your own API key, and the file carries no vendor import.

CriterionRetracioChecksumAI test generator
Follows the conventions of your repositoryYespartial: reads the connected repository and existing tests via repo mirror; generated tests use Checksum fixtures (init() from @checksum-ai/runtime) and .checksum.spec.ts / .checksum.md file pair; custom style guides and integration with existing testing infrastructure listed as Scaling plan features; in IDE integration user points the agent to existing tests as examples
Where tests runLocal or cloudboth: test execution locally or in CI via Checksum CLI (also with Playwright directly); detection, generation and healing run as cloud agent sessions
Who owns the test codePlain test files in your repositoryopen code in user's repo: Playwright test files delivered as pull requests to the user's repository; tests import Checksum fixtures, replaceable with standard Playwright imports
Works offlinefalse for generation and healing: the agent runs in the cloud; CLI config requires apiKey; local test runs keep reports local unless hostReports is enabled
Maintenance modelEdit the test like any other fileself-healing: auto-recovery during execution (CLI AI fallback) and auto-healing after CI failures (agent fixes test code and opens a PR for review)
Output frameworksPlaywrightplaywright

Alternatives checked on 2026-09-08 against their own websites and documentation. Empty cell: not confirmed.

Checksum fits when…

  • You want a cloud agent that generates and verifies tests autonomously in the cloud, and delivers the result as pull requests for review (checked 2026-09-08).
  • Failing tests in CI should be fixed by a healing session that edits the test code and opens a PR, and you accept a repo mirror keeping a synchronized view of your test repository (checked 2026-09-08).
  • A plan priced by the number of workflows Checksum maintains, with a free trial on each plan, matches how your team buys tooling (checked 2026-09-08).

Retracio fits when…

  • Testers already walk through scenarios by hand, and each run should leave behind a spec with assertions, saved into the folder you choose and committed by a person on the team.
  • Every new spec has to import the same fixtures and helpers as the existing tests and pass the repo's lint and formatter config, with no vendor runtime in package.json.
  • Security review requires that source code, cookies, and credentials stay on the machine and that the model is called through an API key the company controls.

Questions

Who owns the AI generated test code?

With Checksum, Playwright test files land in your repository as pull requests. They import Checksum fixtures, such as init() from @checksum-ai/runtime, and run with vanilla Playwright once the Checksum imports are replaced with standard ones (checked 2026-09-08). With Retracio, the file is written directly into your local checkout with no product import, plugin, or runtime, so nothing needs replacing. Uninstall the app and every generated test keeps working in your CI.

Does the generated test follow our coding standards?

Checksum reads your existing tests and codebase through the repo mirror; generated tests use the .checksum.spec.ts and .checksum.md file pair, and custom style guides and integration with existing testing infrastructure are listed as features of a specific plan (checked 2026-09-08). Retracio builds a conventions profile from the tests already in the repo before writing a line: folder pattern, naming, fixtures, assertion style, lint and formatter config. The spec lands next to similar tests and imports the page objects the repo already has.

Do the generated tests break when the UI changes?

Checksum handles this during execution and after it: the CLI includes an AI fallback that attempts to fix failing tests in real time, and after CI failures a healing agent analyzes the failures, fixes the test code, and opens a PR for review (checked 2026-09-08). Retracio works at generation time. It resolves each element to a role, label, or test-id locator, follows the repo's own locator convention, and when a step fails on review regenerates only that step, leaving the rest of the file and your edits untouched.

Do AI generated tests require maintenance?

Both tools produce a Playwright file that lives in your repository, so maintenance means editing that file. Checksum adds a healing agent that runs in the cloud and proposes fixes as pull requests after CI failures (checked 2026-09-08). With Retracio, a change in the application is handled by re-running the manual scenario or by regenerating the single failing step from the review screen; the assertions you marked during the session stay in the file. No product service watches your CI.

Does it work offline?

Checksum's agent runs in the cloud, and the CLI configuration requires an API key. Local test runs keep reports on the machine unless report hosting is enabled (checked 2026-09-08). Retracio records the session, scans the repository, and runs tests on your machine. It needs a network connection only for the generation call to the model through your own API key. The generated tests run offline like any other Playwright spec.

Does my source code leave my machine?

Checksum maintains a synchronized view of your test repository through a repo mirror, reading your existing tests and codebase and writing back by opening pull requests (checked 2026-09-08). Retracio reads the local checkout only and never clones or uploads it. What goes out is the recorded steps, trimmed DOM snapshots of the pages under test, and excerpts of the conventions profile, sent to Anthropic Claude through an API key you provide. Source code, generated tests, cookies, screenshots, and credentials stay local.

How do the generated tests integrate with CI/CD?

Checksum tests execute locally or in CI through the Checksum CLI, or with Playwright directly; reports upload to the dashboard, and CI failures can trigger a healing session (checked 2026-09-08). Retracio writes standard Playwright, so the job that already runs your suite picks the new spec up with no extra step. GitHub Actions and GitLab CI are supported today; Jenkins, Azure Pipelines, and CircleCI are planned.
Join the waitlist