Skip to main content
Collect the details below before you contact us on Discord, open a GitHub issue, or email founders@tracecat.com. We need your Tracecat version, and your Helm chart version if you deploy with Helm, before we can start debugging.

Required information

Tracecat version

Get the version from the running API:
You can also read it from your deployment configuration:
  • Docker Compose: run docker compose images api ui, or check TRACECAT__IMAGE_TAG in your .env file.
  • Kubernetes: run the command below. image.tag and uiImage.tag override the chart’s appVersion, so check the running images rather than the APP VERSION column in helm list.
  • AWS ECS Fargate: check the tracecat_image_tag Terraform variable.

Helm chart version

If you deploy with Helm, send the chart version from the CHART column:
tracecat-0.6.0 means chart version 0.6.0. Also tell us which values you changed from the defaults, with secrets removed.

Latest release

Check the changelog for a fix to your issue, then update to the latest Tracecat image, Helm chart, or Docker Compose files: Tell us whether you updated and whether the issue still happens on the latest release. If you can’t update yet, say so and include the version you plan to move to.

Details that speed up debugging

  • Exact error: the full message and where it appeared (UI, workflow run event, agent chat, API response, or service log). Add a screenshot for UI issues.
  • Time: when the issue happened, with timezone, so we can match it to logs.
  • Reproduction steps: the smallest workflow, agent preset, or API call that reproduces the issue. Export the workflow definition as YAML and remove secrets.
  • Regression: whether this worked on an earlier Tracecat version, and which one.
  • Recent changes: upgrades, environment variable or Helm value changes, infrastructure, proxy, or network changes made before the issue started.
  • Scope: whether the issue affects every workflow, user, and workspace, or only some.
  • Service health: output of docker compose ps or kubectl get pods -n tracecat, including restart counts.
  • Logs: logs from api, worker, executor, agent-worker, and agent-executor around the time of the issue. For UI issues, add errors from the browser console.
  • Deployment setup: your execution backend (TRACECAT__EXECUTOR_BACKEND), Temporal (bundled, self-hosted, or Temporal Cloud), PostgreSQL (bundled or external), authentication type, and whether you run behind a reverse proxy, a custom CA, or in an air-gapped environment.
Remove secrets, API keys, tokens, and customer data from everything you send. Never share your .env file or TRACECAT__DB_ENCRYPTION_KEY.