AI reverse debugging · Python & C

Debug backwards.
Ship forward.

X-Debug traces every bug to its root cause with a 3-layer engine, visualizes cause and effect as a live causal graph, and stages ranked AI fixes in a safe git-branch sandbox — so nothing changes until you commit.

causal_trace.py
# eval() on untrusted input
- config = eval(data)
+ config = json.loads(data)
# root cause → data → eval → RCE

The X-Debug engine

Three cooperating layers turn raw code into an actionable diagnosis.

Layer 1

Expert System

Rule-based static analysis catches syntax and structural anomalies — unused variables, null pointers, uninitialized memory, and insecure input.

Layer 2

ML Analysis

A CodeBERT-style model assigns a defect probability to your file and every finding, so you fix what actually matters first.

Layer 3

NLP Explanation

Every bug gets a plain-English narrative — what it is, why it's dangerous, and how the fix resolves it.

Built for the hard bugs

Not just detection — a full workflow from root cause to a safe commit.

Causal Graph Debugger

A live visual graph of cause-and-effect between variables, functions and control flow — hover any node to trace how a bug propagates.

Multi-Fix Ranking Engine

Multiple alternative fixes, each scored by confidence, efficiency and code quality, ranked best-first.

Git-Branch Sandbox

Fixes are staged in a safe sandbox with a red/green diff viewer. Nothing merges into your live code until you hit Commit.

Find the bug. Understand the cause.

Sign in and run your first analysis in seconds. Works offline as an installable app.

Get started free