Building agentnative, the agent-native CLI standard. AI-native operator of a 300ドルM construction company. Former Uber marketplace operator, two patents. Aspiring compounder.
The CLI standard for AI agents. Paste a CLI at anc.dev, get a scorecard, share. Eight RFC 2119 principles, four artifacts:
- spec defines eight principles with machine-readable requirement IDs.
anclints CLIs against the spec.brew install brettdavies/tap/agentnative.- anc.dev/scorecards ranks top CLIs against the spec on a live leaderboard.
- skill bundle teaches Claude Code / Codex / Cursor to invoke
ancand remediate findings.
- bird and xurl-rs are agent-native CLIs for the X API that dogfood the standard.
- Streams Group is a 300ドルM construction company, AI-native from the ground up.
Five years at Uber marketplace ops, scaling from local to global: Manager (San Diego, 45 US-West markets) → Senior Manager, Regional (LA, 150ドルM incentive & promotions P&L, 30-person org) → Senior Manager, Country (Melbourne) → Lead Engineer, Global Workforce Systems (JUMP, 25 markets). Founded regional-ops functions that were later adopted as global best practices; led internal org migrations with measurable cost and headcount efficiency across rideshare and micromobility.
What I'm proudest of is on the driver side. Invented the 3-strikes driver quality policy. The patents (USPTO 9,949,088 and 9,940,654, Network System with Scheduled Breaks) came out of fixing time-to-first-trip for drivers returning from a gap. Drove Melbourne's dynamic-pricing and service-zone work: +20ドルM/yr gross bookings, +15% driver retention. Built JUMP's four-pillar workforce optimization system in Python/PostgreSQL/BigQuery: 125ドルM annual staffing cost reduction.
After Uber: Lime (−75% operational cost across 100+ markets), Gopuff (17ドルM loss mitigation via fraud detection, ×ばつ SKU-scraper rewrite). Gauntlet AI Fellow, 2025 (0.8% acceptance).
I work comfortably across business and engineering, with full end-to-end ownership from spec to ship to operations.
The code speaks for itself.