Typescript

git-scenarios came together fast because the project it serves, Coco’s TUI git workstation, had very clear testing needs. A couple of weeks after the v0.1.0 extraction it’s now stable at v1.0.0: 32 scenarios, 60+ atoms, five framework adapters, 527 tests across three operating systems. The v1 work wasn’t about features, it was about trust: typed error classes, precondition guards, property-tested determinism, cross-platform CI, and shedding the last of its coco branding.

I have written the same 80 lines of git test setup more times than I want to count. @gfargo/git-scenarios v0.5.0 extracts the helper layer that grew up inside Coco: a registry of named scenarios (mid-merge-conflict, submodule-with-history, mid-bisect, and 8 more), a composable atom API for building your own, and a tool-agnostic CLI that drops any tool into a real repo in a known state.

Doorman 2.0 ships Cloudflare WAF support alongside Vercel Firewall, a new .doorman.json config format, and 1,100+ tests backing a production-ready multi-provider firewall CLI.

A look at everything that’s landed in ink-enhanced-select-input on the road to 1.0: a headless hook for custom renderers, multi-select mode, inline search, item groups, scroll indicators, Escape/cancel support, and 133 tests – including a real bug caught by writing edge-case tests for backspace handling in searchable mode.

After building Spec Shaver as a CLI for trimming oversized OpenAPI specs, I kept finding myself wanting a visual way to do the same thing. SpecForge is what came out of that itch: a web app for importing, curating, and exporting OpenAPI specs in the browser.

Hitting OpenAI’s 1MB / 30-operation cap on Custom GPT actions when your real API has 150+ endpoints is maddening. Spec Shaver is the CLI I built to reduce a bloated OpenAPI spec down to the operations that actually matter, while keeping the schema valid.

Fargo Flags is the feature-flag toolkit I wished existed for Next.js: zero infrastructure, type-safe, and small enough to drop into any project. Here’s the design call that shaped it and the trade-off I’m still not sure about.

Why I rolled my own SVG path editor instead of reaching for an existing one – what the off-the-shelf options got wrong for my use case, and what I learned about Bezier curves trying to make the handles feel right.

Coco is the AI-powered Git assistant I built because I was tired of writing terrible commit messages at 11pm. The story of how it grew from a one-off script into something I now use daily across every project.

A Sunday backlog on coco that should have eaten the whole day took about an hour. Not because the AI wrote my code, but because steering files and specs meant it already knew the project instead of me re-explaining it every session. Here is what is in mine, and what I left out.