OSS AI search XERJ hits 2,000 daily installs in 3 months, GitHub Rust #1

alexcovo_eth · x · 2026-10-18

Open-source enterprise AI search project XERJ reached 2,000 unique installs per day (760 clones + 1,342 binary downloads) three months after first release, hitting GitHub's Top 10 and holding #1 in the Rust category for 3 days. The author stresses that OSS has zero adoption friction and scales without marketing budget.

Core design principle: build for AI, not people. AI agents are the real customers — picky, highly educated, far more engaged, and they choose products differently; new software shouldn't have AI using grep from 1973 or regex from 1956. The companion XERJ Corpus Hub is a reference-memory platform for agents: it pins a field's open knowledge (docs, code, rules, research) into searchable memory corpora — 511 live corpora across 11 domains, 78.6 GB of pinned text covering security (exploit PoCs, vuln databases), DB, crypto, web, law, health — so any agent that loads one works like a domain expert instead of guessing from training.

Original post →

More from coding & agent

coding & agent channel →