Skip to main content

​
The Problem

Post-incident investigations are slow, manual, and depend on institutional knowledge that walks out the door. Root cause analysis is inconsistent. Recurring incidents share the same root cause with different symptoms.

​
What It Does

Feed it a production incident. Returns:
  • 5-Why trace
  • Impacted symbol and dependency graph
  • Historical pattern matches against prior incidents
  • Structured remediation brief

​
Try It

activate nablr
Set agent to forensic_engineer. [describe the incident]
Always provide context upfront — the more specific, the faster the RCA: 
Set agent to forensic_engineer. The payment API returns 500 errors when the cart contains more than 10 items. Started after yesterday's deploy.
Then proceed to fix: 
Set agent to strategist
Set agent to developer. Implement the approved fix

​
Expected Output

  • nablr-reports/forensic_engineer/rca_*.md — 5-Why trace + root cause
  • Impacted symbol graph
  • Remediation brief with recommended fix
  • Regression test to prevent recurrence

​
Who This Is For

  • SRE and platform teams
  • On-call engineers
  • Engineering managers running post-mortems
  • Any organization where incident resolution time directly affects revenue or SLA
Paste the error message, stack trace, or log snippet directly into the prompt. ForensicsIQ uses concrete evidence, not guesses.