AI Workflow OS — AI skill for Claude Code
Append-only ledger + 5-state machine that keeps 6 divisions honest across 5 AI platforms (Claude, ChatGPT, Gemini).
How to install AI Workflow OS
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open bangnevgo/ai-workflow-os and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What AI Workflow OS does
Append-only ledger + 5-state machine that keeps 6 divisions honest across 5 AI platforms (Claude, ChatGPT, Gemini). CLI for LLM agent workflow tracking: evidence-gated 'done', dependency holds, tamper-evident JSONL history, and Memory & Dreaming (Karpathy wiki pattern) cross-session memory for AI agents. Zero-dependency Python.
Alternatives in AI
- UI UX Pro Max Skill — An AI SKILL that provide design intelligence for building professional UI/UX multiple platforms 45.8k ★
- Installation Guide - Claude Skills Library — Complete installation guide for all 170 production-ready skills across multiple AI agents and platforms 5.3k ★
- Surface — Show ranked attack surface for a target from its recon manifest + hunt memory 3.8k ★
README
🤖 ai-workflow-os
One append-only ledger. Six business divisions. Five AI platforms. **Zero ops team.**
**TL;DR — keeping multiple AI agents in sync.** Running several LLM platforms (Claude, ChatGPT, Gemini, …) on the same work? This is a **single source of truth for AI agent teams**: an append-only, tamper-evident JSONL ledger where "done" is **evidence-gated**, closed projects are held by **open dependencies**, and cross-session memory is compiled from session logs using the **Memory & Dreaming / Karpathy wiki pattern** — so agents never overwrite each other's finished work.
You run Content, Marketing, Operations, Sales, Finance, and Website through **five different LLM platforms** — no ops team, just you and a fleet of agents. The hardest part is never the models. It's that every platform carries its *own* memory and its *own* interpretation of what happened.
Platform A says the page is live. Platform B, two days later, rebuilds that same page — it never knew the work was done. Platform C logs another division's work as its own, because it was only "helping".
**ai-workflow-os is the single source of truth they all have to agree on.** Agents may claim anything in
Related Skills
Canny
Stops AI coding agents from claiming work is done without evidence. Deterministic hooks decide, TypeSafe's Jev
Cobol Legacy Ledger
Learn COBOL through a live banking system — 18 programs, 6-node settlement network, real-time simulation, Pyth
Self Growth Loop
Self-governed skill-growth loop for AI agents: propose -> owner-approve -> trial -> council-review -> adopt, w
Nissth
Deterministic execution framework for LLM coding agents. Pre-structured context, append-only state, no explora
AI Loop Skill
Your agent says "done" and it is not. A gated multi-agent circuit for Claude Code: whoever implements does not
Sherwanis Claude Md
Drop-in AGENTS.md / CLAUDE.md that keeps AI coding agents honest: ask before assuming, keep it simple, change
Related Agents
Continuity Supervisor
Cinema Studio continuity supervisor and QA scorer. Keeps the continuity ledger, shot ID tracking, the realism
Advisor SEO Aeo
SEO/AEO expert for Altis — getting paywalled venture research cited by AI answer engines (ChatGPT, Perplexity,
Hva Desk
The HVA Desk — reads Meta insights by impression volume, runs the deterministic CLEAR scorer, and converts the