HELP
Troubleshooting
A repeatable checklist for connection, access and workflow issues.
Overview
Diagnose issues from observable evidence. Start with the narrowest failing boundary and avoid sharing credentials, personal data or confidential payloads in diagnostic material.
Confirm the basics
Check the environment, URL, account, role, time of failure and whether the issue is reproducible.
Check access
Confirm the user has access to the workspace, source and action. Review recent permission or policy changes.
Inspect connection health
Check connector status, credential expiry, network reachability, rate limits and service status.
Inspect the workflow
Identify the first failed step, input validation error, timeout or denied action. Compare expected and actual behaviour.
Capture safe evidence
Record timestamps, correlation IDs, redacted error messages and minimal reproduction steps. Remove tokens and sensitive payloads.
Escalate with context
Include impact, frequency, environment, recent changes and what has already been tested. Never include passwords or API keys.
Worked example
An integration step fails
Diagnose the first failing boundary and use evidence to choose the next step, rather than repeatedly retrying an unknown error.
- 1Record the time, environment, workflow step and safe correlation identifier.
- 2Check the connector status, identity, permission scope and credential validity.
- 3Inspect the redacted error category: denial, timeout, throttling or validation failure.
- 4Retry only when the cause is understood and the operation is safe to repeat.
- 5Verify the final state in the target system and record the resolution for future operators.
Was this guide useful?
Send feedback or flag a gap. Include the article link so the team can improve the right guide.