Z: Local-first search layer for humans and agents

Look, I've been covering open source long enough to see a pattern: every couple of years, someone reinvents grep and calls it a revolution. But Z is different—it's not just searching your files, it's searching them *semantically*, and it's doing it locally. That's the part that matters. We've spent a decade shipping our data to the cloud and begging for it back, and the pendulum is finally swinging home. Local-first isn't a buzzword anymore; it's a survival strategy. The "for humans and agents" bit is the real story here. We're entering a world where your terminal isn't just for you anymore—your AI copilot is in there too, rummaging through your codebase, your notes, your config files. And it needs a search layer that understands *meaning*, not just regex patterns. Z is positioning itself as the Rosetta Stone between human intuition and machine pattern-matching. That's not a gimmick; that's a necessity. But let's not get carried away. The HN thread has exactly one comment, and the points are in the single digits. This is a project that's early, unproven, and competing in a crowded space—there's ripgrep for speed, everything for file search, and a dozen vector databases fighting for the semantic crown. Z's bet is that it can fuse those worlds into one tool that doesn't require you to stand up a Postgres cluster just to find a damn file. The deeper question is whether local-first can survive the agent economy. Every major AI company wants you in their cloud, using their search, paying their API bills. A tool that keeps everything on your machine is a threat to that business model, which means it's either going to get acquired, ignored, or become the backbone of a decentralized agent protocol. My money's on the latter—the open source community has a way of rallying around anything that tells the cloud to go pound sand. Source: https://github.com/zvec-ai/zvec-grep