Project Open source · MIT · VS Code extension
AI Preflight
See your prompt's risk level, token budget, context waste, and security exposures before you send it to your AI tool. Live in the editor, with a @preflight chat helper.
Projects
Tools I've built — mostly open source.
AI Preflight — a project family
Security and hygiene for the AI-agent era — before the prompt, before the PR.
As AI coding agents gain more autonomy — reading your files, following instruction files, calling MCP tools — the risky moment moves upstream, to what enters the model's context and what the agent is allowed to do. AI Preflight is a small family of tools built for exactly that moment: one detection engine, two surfaces. A VS Code extension that analyzes your context live before you prompt, and a CLI that gates your changes before you open the PR. Everything runs locally and is detect-and-explain only — nothing is modified, nothing is sent anywhere.
Project Open source · MIT · CLI
vet
A pre-PR security gate for AI-assisted code. Scans for exposed secrets, trap files, and risky agent/tool configs. One command, local, CI-ready.
Project Open source · MIT · npm package
@deepwissen/preflight-core
The shared, zero-dependency TypeScript engine powering both tools. Secret, trap-file, instruction-integrity, and MCP-config scanners you can build on.