AI coding assistants are powerful but unpredictable. Without guardrails, they skip tests, hallucinate patterns, and create technical debt faster than any human. Samuel is an open-source AI development framework that gives professional teams 35+ testable rules, 21 language guides, 33 framework guides, and 24 workflows — so AI assistants ship consistent, production-grade code every time. Cross-tool compatible with Claude Code, Cursor, Codex, and Copilot.

Samuel grew out of a real problem: our engineering team at Cuemby was using AI coding assistants across multiple tools and getting inconsistent results. We needed something opinionated, portable, and enforceable — not vague suggestions, but testable standards. Samuel is the framework we built to solve that. It's fully open-source, AGENTS.md compatible, and already used across teams building production software. Version 2.0 ships with an autonomous AI coding loop (samuel auto), per-folder instruction hierarchies, and a CLI with 14 commands for component discovery and management. This session walks through why we built it, how it works, and how any engineering team can adopt it — regardless of which AI tools they're using today.