
Fuad P.
Python Code Auditor and AI Project Fixer
Skills

Bekijk mijn diensten


Portfolio
Werkervaring
Author of Plumblint, prompt injection detector
Self-Employed / Freelancer • ZZP
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.