Vinix24/vnx-orchestration

VNX

VNX runs AI coding CLI workers in tmux, isolated git worktrees, through review gates, with an append-only NDJSON receipt per dispatch.

It is a local control plane for the AI coding CLIs that already sit on your machine. One orchestrator dispatches work to ephemeral workers; each worker runs in its own git worktree; review gates decide what merges; every dispatch leaves a receipt. VNX drives claude, codex, gemini, kimi, and local ollama with no vendor SDK. It calls the CLIs as subprocesses and never imports a provider library.

Most agent projects build SDK-native agents. I orchestrate the binaries instead. The difference shows up in the audit trail: I can reconstruct what was dispatched, what was reviewed, what merged, and what each gate cost.

I built this for my own work, across 2,000+ hours of Claude Code and 1,450+ tests. It is open source because the architecture is portable. Source is at github.com/Vinix24/vnx-orchestration.

This is not a security sandbox; it isolates work with tmux sessions and git worktrees. It is not compliance certification; it produces a local, append-only, inspectable audit trail. It is optimized for human-gated coding workflows, not fully autonomous merges.

What's new in 1.0

• pip-installable. pip install vnx-orchestration, then vnx init, vnx migrate, vnx doctor. No repo clone required to scaffold a governed project.
• Provider-agnostic skill injection. One skill folder, one structured prompt (role, assignment, resource index), identical for claude, kimi, codex, and deepseek workers. ADR-022.
• Realistic benchmark methodology. A field-tests harness that measures provider lanes on production-derived tasks with programmatic verification per task, an LLM-judge fallback, and cost per quality-point. It lives in the repo under scripts/benchmark/field-tests/ (not the pip package