Overview
Trodo continuously scans your runs, spans, and UX events for known failure patterns, groups recurring ones into ranked issues, and can fix them with a coding agent.
A single failed run is noise. The same failure across hundreds of runs is an issue. Trodo watches your traces and product events for known failure patterns, groups recurring ones, ranks them by severity, explains the likely cause, and can hand the fix to a coding agent with Heal.
What it does
Detect
Dozens of built-in detectors scan runs, spans, and UX events for known failure patterns.
Rank
Each issue gets a severity score from reach, frequency, growth, and business impact.
Explain
An analysis names what's happening, the likely cause, and how to fix it, with evidence.
Fix
Send an issue to a coding agent with Heal and get a branch back.
How it works
Instrument
Send runs and spans (and browser events for UX issues) with the SDK.
Detect
Detectors run automatically and group recurring failures past a threshold into issues.
Triage & fix
Read the analysis, set status, and optionally Heal it.
Turning it on
Issues reads the data you already capture. There is nothing to configure to start: instrument your agent with Observability so runs and spans flow in (and add the browser SDK if you want UX issues like rage clicks and visible errors), and detection runs automatically. See Detectors for what's caught and how an issue is formed and scored.
Where issues show up
- The Issues list, grouped by category, filterable by status.
- An issue's detail: impact tiles (affected users, occurrences, severity, 7-day trend), "what's happening", ranked likely causes, "how to fix" with an owner and effort hint, and the evidence (the exact runs, spans, UX signals, and errors behind it).
- On a run or span: a chip linking to the issues it belongs to.
Lifecycle
An issue moves through open → acknowledged → resolved, and you can mute ones you don't care about. If a resolved issue starts spiking again, its regression flag brings it back to your attention.