Skip to main content
Insight

AI root-cause copilot for n8n

Stop reading raw execution JSON by hand.

Paste a failed n8n execution — your own instance and API key, or just an exported JSON file — and Insight tells you which node broke, why, and how confident it is, in seconds.

Diagnose a failure
No account required
Paste an execution ID plus your own instance details, or upload exported execution JSON — either way works standalone.
A thin, honest frontend
All diagnosis logic runs in an n8n workflow. This page only forwards your request there and renders whatever comes back.
Calibrated, not overconfident
Low-confidence diagnoses are shown visibly hedged, never presented with the same certainty as a high-confidence result.

How it works

Two ways in: paste one failure with no setup, or connect an instance so Insight watches it going forward.

  1. 1

    Connect your instance

    Sign in, then give Insight your n8n instance's base URL and an API key. Nothing on your instance changes yet.

  2. 2

    See what's unprotected

    Insight scans the instance and lists every workflow on it, flagging which ones already have failure detection wired up and which don't.

  3. 3

    Click Add workflow

    Insight creates and activates its error-workflow template on your instance and points that workflow's Error Workflow setting at it — automatically. It never edits that workflow's own nodes or connections.

  4. 4

    Get diagnosed, not just alerted

    The next time that workflow fails, Insight fetches the full execution, redacts secrets, and produces a plain-English root-cause diagnosis — pushed to Slack and logged on your dashboard.