The same root cause keeps coming back because nobody tracks it. I built a zero-dep CLI that does.
The article describes a common incident management problem: postmortems are written, action items filed, but the same root cause can resurface months later because the fix never shipped and no one remembers the first incident. An SRE thread mentioned using Rootly to automatically track root causes, but that is a paid, hosted feature of a full incident-management platform. Most teams only have a folder of markdown postmortems that are not compared against each other. To address this, the author built rootecho, a zero-dependency CLI that flags when a new incident's root cause echoes a past one and shows whether that past incident's action items were completed. Each postmortem is stored as a JSON record containing a free-text root_cause and/or curated root_cause_tags, plus action_items with a status. An example record is provided: {"id": "INC-2026-014", "title": "Payment webhook retries exhausted", "root_cause": "webhook retry queue misconfigured to drop after 3 attempts, no dead"}.
Helps teams prevent recurring outages by tracking root causes and action item completion.