AI tooling engineer. I build open-source tools that help AI agents write code that actually works.
Every project here addresses the same gap: AI-generated code looks right but breaks at runtime.
My tools make it checkable — before it reaches production.
An Agent Skill that stops AI-written code from being built on APIs that do not exist.
Six validation gates — from Python builtins to npm, PyPI, and Docker Hub — with a severity model, SARIF output, and CI integration.
Contract-drift monitoring for MCP servers and AI connectors.
Snapshot, diff, judge, report — deterministic and reproducible. Catches breaking changes before they break your integration.
Generate a compressed, AI-ready context pack for any codebase.
Architecture, endpoints (20+ frameworks), dependencies, TODOs, security findings, conventions — as Markdown, JSON, or a paste-ready prompt. Zero dependencies.
Python Git GitHub Actions CLI tools REST APIs LLM / AI agents Prompt engineering Static analysis Security scanning
- Every public repo ships with tests, documentation, and CI.
- Images are rendered and inspected before they are committed — a broken image is the first thing a visitor sees.
- Version tags are immutable; releases include changelogs.
Ideas, issues, and PRs welcome.