pattern-surgeon
Problem
Agent-driven refactoring is trusted rather than verified, so a bad pattern choice lands silently.
Approach
Score every candidate, reject poor fits with a stated reason, and revert automatically when tests fail.
Outcome
Qualitative — determinism over trust
9 languages · deterministic rollback
Ask an agent to apply a design pattern and it will, whether or not the pattern fits. pattern-surgeon scores each candidate first and refuses the poor ones out loud, with a reason. If the tests go red after a refactor, it reverts itself. Nine languages, and the interesting property is not the breadth — it is that the failure mode is "nothing happened" rather than "something subtly wrong happened".
What I did
- Refactoring agent across 9 languages: scores each candidate, refuses poor fits with a stated reason, and reverts itself when tests fail.
More work
AI ToolingAgentic Development Life Cycle (ADLC)
The engineering manifesto that replaced ad-hoc ticket work at my company with a phase-gated, human-approved delivery pipeline.
AI ToolingAgentic Delivery Harness
A delivery harness that drives an approved requirements document to shipped code through phase gates, and runs the same way on any coding agent.
SecurityCyberWise
A security awareness and phishing simulation platform that banks run against their own staff to find out who is at risk, and then train them.
Building something that needs this kind of engineer?
Hiring, contract, or an architecture problem — tell me the shape of it and I’ll come back with something specific.