s
steel_loop

Fuad P.

@steel_loop

Python Code Auditor and AI Project Fixer

Azerbaigian
Inglese, Azero, Russo, Turco
Alcune informazioni sono riportate in lingua inglese.
Chi sono
I read code and find what is wrong. Python, JavaScript, TypeScript, Go, Java, C, C++, Rust, SQL, shell, Dockerfiles, CI and dependencies. Python goes deepest: data is traced from input to the dangerous place. Injection paths, bypassable access checks, credentials in the source, crashes on bad input, unhandled cases, lost updates between concurrent callers, memory and time with no ceiling, vulnerable dependencies. Every finding names the file, the line and how to see it. Same evidence, byte-identical report. Every file examined and skipped is listed. I also fix broken AI-built apps.... Continua a leggere

Competenze

s
steel_loop
Fuad P.
offline • 
Tempo di risposta medio: 1 ora

Consulta i miei servizi

Risoluzione dei problemi e miglioramenti
I will fix your broken lovable bolt new v0 or replit app
revisione codice
I will review your vibe coded or python app for security and logic bugs

Portfolio

Esperienza lavorativa

Self-Employed_/ Freelancer

Author of Plumblint, prompt injection detector

Self-Employed / Freelancer • Lavoratore autonomo

Sep 2026 - Present • 2 mos

I build and publish developer tools in Python, and I audit Python codebases for defects. Plumblint - a prompt injection detector for LLM applications. Published on PyPI (pypi.org/project/plumblint) and GitHub (github.com/knbtz65v2c-pixel/plumblint). Python 3.11+, zero dependencies, no network calls, no telemetry. Source-available under PolyForm Noncommercial 1.0.0. The release ships its own calibration numbers together with a written section on what the detector does NOT catch, and states that the evaluation corpus was written in-house, so the published numbers are an upper bound rather than a guarantee. I keep the same standard in client work: every finding comes with the command and the output that produced it, and every report says plainly what was left unchecked.