e-volv

Planning playbook

AutoRootCauseAnalysis

Investigates bug-labeled issues, traces the root cause in code, and posts the analysis with a fix.

When an issue is labeled 'bug', investigate the codebase, trace the reported behavior to its likely root cause, and post an analysis comment with a suggested fix and a confidence rating.

Identifier
auto-root-cause-analysis
Version
1.1.1
Steps
4
Triggers
1
  • issues
  • debugging
  • analysis
  • automated

When it runs

Issueissue.labeled
Settle delay
10s after the event

Only when all hold

  • issue.addedLabel equals bug
  • actor.username not contains [bot]

The pipeline

The graph below is the one the workflow opens with in the builder — same steps, same layout, drawn on the same canvas. The run playing through it is a simulation; the branches and conditions are real.

React Flow
  1. 01
    Issue Labeled bugtrigger

    The event that starts the run.

  2. 02
    Clone Repositorygit.clone

    Shallow-clones the repository at the right ref.

  3. 03
    Analyze Root Causeagent.run

    One agent works the task with its toolkit.

  4. 04
    Publish Root Cause Analysisreport.publish

    Posts or updates a deduplicated report comment.

The agent

Root Cause Analyst

Base type
Senior Developer
Temperature
0.2
Max iterations
35
Tools
7

Filesystem · 2

  • read_fileRead File · read
  • list_dirList Directory · read

Git provider · 2

  • list_issuesList Issues · read
  • add_labelsAdd Labels · write

Code search · 1

  • code_searchCode Search · read

Terminal · 1

  • run_terminal_cmdRun Terminal Command · write

Status · 1

  • update_statusUpdate Status · write