Help with a test or integration
Troubleshoot access, missing evidence, queued runs and integration setup, and send Testsy support a useful report without credentials.
Start with the result you have#
A plan, a queued job and a final result are different stages. Open the test's run page and check its current status before starting another run.
| Situation | First check |
|---|---|
| The plan asks questions | Provide the missing expected behavior, test account or starting state, then review the plan. |
| The run is queued | Read the queue reason: available capacity, workspace concurrency or runner availability. |
| The test cannot start | Check the returned reason, your workspace allowance and whether its requested services are available on your plan. |
| A check failed | Compare the expected and observed values and inspect the screenshot or replay at that step. |
| The run is blocked | Read whether the obstacle is site access or a tester limitation; resolve it before relying on this journey. |
| The run is inconclusive | Inspect what the tester could not establish. Clarify an ambiguous check or investigate the missing evidence. |
| The run ended in error | Keep the error and run ID for investigation. An infrastructure problem is not proof of an application defect. |
See the complete verdict guide and API errors and limits. A legitimate failed assertion must stay a failure until the application or the agreed requirement is corrected.
Check access and test data#
The hosted browser must be able to reach the start address. A private network or your computer's localhost is not automatically accessible. Use an authorized staging deployment and review the network guide for the supported setup.
For account flows, add a dedicated sign-in account to the project and name it in the test. Check that it still works and has the right role. Use synthetic tasks, documents and member accounts so that repeat tests do not disturb customers.
For signup verification, enable the run mailbox and inspect the Messages tab. Do not assume a message arrived because the application displayed a delivery notice.
Check an agent or CI connection#
Confirm that the client uses the intended Testsy workspace and an active connection. For OAuth, review the connected app and its permissions in Account settings → Connected apps; reconnect if its grant expired or was revoked. For API keys, check the key is active and keep it in the client's environment or secret store. Running saved tests needs run permission; creating and editing tests needs write permission.
Use the current MCP setup or GitHub Actions example. A successful connection or tool discovery does not mean a test has run: wait for the final run result.
If a CI job succeeded despite a blocked or inconclusive run, inspect its failure policy. These verdicts are not passes.
Contact support with the evidence#
Email contact@testsy.dev or use Something wrong? Tell us on a documentation page. Include:
- The run ID, or the test/project ID if the run could not start.
- What you expected, what happened, and the approximate time and timezone.
- The exact error code and client name for an integration problem.
- A screenshot or a read-only result link when appropriate.
Do not include API keys, passwords, session cookies, live verification links or customer data. Check a result and its evidence before creating a share link. Support does not need your API key to identify a run.
There is no published response-time guarantee. If the app itself is inaccessible, include the address and visible error in your email.