I build practical, explainable tools for engineers and developers — from cross-checking quality documents to auditing repositories and coordinating AI-assisted projects.
My background is in automotive supplier quality, root cause investigation and corrective actions. I apply the same discipline to software: explicit rules, traceable findings, local-first workflows and human approval gates.
Explore my projects · How I work · Get in touch
Quality Docs Validator — spot potential PFMEA ↔ Control Plan gaps
For quality, process and manufacturing engineers. Check two Excel workbooks for missing controls, unmatched process steps and other potential inconsistencies before technical review.
| Example run on bundled synthetic files | Result |
|---|---|
| Potential findings | 3 critical · 3 warnings |
| Severity-weighted score | 43/100 — FAIL |
| Outputs | Terminal, Markdown, JSON |
The example is deliberately seeded with inconsistencies. The tool highlights potential issues for engineering judgement; it does not certify compliance or replace review.
git clone https://github.com/migmcc/quality-docs-validator.git
cd quality-docs-validator
pip install -e .
qdv pfmea-control-plan --pfmea examples/pfmea.xlsx --control-plan examples/control-plan.xlsx --out report.mdSee the full example report · Read the six checks · View source
| Project | What problem it solves | Start here |
|---|---|---|
| Public Repo Audit | Checks whether a local repository has the documentation, tests, CI and packaging signals needed for public release. | README |
| Agent Skill Quality Gate | Deterministic quality checks for SKILL.md files, including basic safety heuristics and CI-friendly output. |
README |
| ProjectPilot | Keeps track of the project lifecycle, required artefacts and human approval gates, without executing agents. | README |
| Agent Desk | Structures and coordinates manual AI-assisted work across Codex, Claude Code and local models. | README |
Browse all public repositories →
- Useful before flashy: start with a concrete problem and a runnable example.
- Evidence over claims: link to documentation, real outputs and tests.
- Human in control: software assists decisions rather than hiding them.
- Transparent scope: state what each project does not support.
- Local-first when practical: prioritise data control and understandable execution.
My professional work includes supplier claims and warranty analysis, 8D / SCAR, IATF 16949, ISO 9001, AIAG Core Tools and VDA methods.
I also contribute to work on reusable AI skills and agent workflows. Collaborative projects are distinguished from the tools I maintain in this portfolio.
Based in Braga, Portugal.
Project capabilities and example outputs are drawn from their linked repositories. No traffic or conversion claims are implied.