Debugging Agent

What Debugging Agent does

A debugging agent treats a failure as evidence. It reads the error, gathers logs or traces, forms a next check, and only then edits. It is not allowed to randomly rewrite the file that 'looks suspicious'.

Error in, reproduce if possible, inspect, hypothesize a check, run the check, then patch. Conceptual simulation.

Workflow

Each step maps to a node in the lab above. Scroll up to play and watch the packets travel.

1. 01

   Input

   Receive the failure

   Load stack trace or failing test.

2. 02

   Plan

   Pick the next check

   Choose an observation.

3. 03

   Tool

   Reproduce

   Run the failing test or command.

   Failure path

   If you cannot reproduce, say so. Do not patch at random.

4. 04

   Tool