ariffazil/arifOS

Independent policy kernel for AI agents: judge before execute. F1–F13 floors, VAULT999 receipts. Not a model. Not a chatbot.

Latest release v2026.10.17 — 1!2026.10.17 — discovery compatibility + registry-grade tool definitions · 17 Oct 2026

10 files for Claude Code, GitHub Copilot, Codex, OpenCode and Cursor: arifOS Governor, Deploy Gate, AAA Judge, Runtime Triage and 6 more — 7,715 tokens loaded in every session.

Open on GitHub arifos.arif-fazil.com

Instructions

arifOS copilot-instructions.md .github/copilot-instructions.md C 5,399 tok
arifOS aaa-a2a-arifos.instructions.md .github/instructions/aaa-a2a-arifos.instructions.md A 638 tok
arifOS AGENTS.md AGENTS.md B 1,217 tok
arifOS CLAUDE.md CLAUDE.md A 8 tok

Cursor rules

cursorrules .cursorrules A 453 tok

Agents

arifOS Governor .github/agents/arifos.agent.md A 58 tok
Deploy Gate .github/agents/deploy-gate.agent.md A 33 tok
AAA Judge .github/agents/judge.agent.md A 39 tok
Runtime Triage .github/agents/runtime-triage.agent.md A 28 tok
Vault Audit .github/agents/vault-audit.agent.md A 30 tok

These files are ariffazil/arifOS's own configuration — they tell Claude Code, GitHub Copilot, Codex, OpenCode and Cursor how to work on this repository, so they are not mods to install elsewhere. Copy one as a starting point and replace the parts that are about this project.