https://github.com/frankxai/agentic-operating-system-standard
Public standard for agentic operating systems: modules, agents, skills, workflows, loops, ledgers, gates, and repo-backed governance.
https://github.com/frankxai/agentic-operating-system-standard
acos agentic-os agentic-workflows ai-agents business-os creator-os investor-os sis standards starlight
Last synced: 9 days ago
JSON representation
Public standard for agentic operating systems: modules, agents, skills, workflows, loops, ledgers, gates, and repo-backed governance.
- Host: GitHub
- URL: https://github.com/frankxai/agentic-operating-system-standard
- Owner: frankxai
- License: mit
- Created: 2026-06-21T21:00:48.000Z (25 days ago)
- Default Branch: main
- Last Pushed: 2026-06-21T21:14:20.000Z (25 days ago)
- Last Synced: 2026-06-21T23:05:37.222Z (25 days ago)
- Topics: acos, agentic-os, agentic-workflows, ai-agents, business-os, creator-os, investor-os, sis, standards, starlight
- Language: Python
- Size: 44.9 KB
- Stars: 1
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE.md
- Security: SECURITY.md
- Agents: AGENTS.md
Awesome Lists containing this project
README
# Agentic Operating System Standard
**A public standard for building agent-native operating systems: Creator OS, Investor OS, Business OS, Domain Intelligence, and future intelligent systems that share one architecture.**
[](https://github.com/frankxai/agentic-operating-system-standard/actions/workflows/validate.yml)
[](./STANDARD.md)
[](./registry/agentic-operating-system-standard.json)
Most agent projects are collections of prompts, scripts, and scattered context files. An **Agentic Operating System** is different. It is a governed system of agents, skills, workflows, loops, ledgers, and approval gates that turns intent into repeatable execution.
This repository defines the shared language and architecture for that category.
## The Thesis
The next durable software category is not one more chat interface. It is a family of domain operating systems where humans, agents, tools, repos, workflows, and memory cooperate through explicit contracts.
The standard has one job: make those systems portable, inspectable, and good enough to become products.
## System Family
| Layer | Role | Example Repos |
|---|---|---|
| SIS | Intelligence kernel: memory, governance, taxonomy, evaluation, provenance. | `Starlight-Intelligence-System` |
| AIS | Discovery and capability routing: agents, skills, repo metadata, `llms.txt`, MCP. | `Agent-Intelligence-System` |
| ACOS | Creator execution substrate: skills, commands, agents, hooks, adapters. | `agentic-creator-os` |
| Domain Intelligence | First asset-intelligence vertical: names, portfolios, verification, offers. | `domain-intelligence-system` |
| Business OS | Founder, revenue, offer, ops, customer, partner, and execution system. | `agentic-business-os` |
| Investor OS | Deal flow, thesis, memo, diligence, portfolio support, and risk governance. | `agentic-investor-os` |
| Asset Intelligence | Cross-vertical asset layer for domains, content, offers, workflows, agents, and ops. | proposed umbrella |
```mermaid
flowchart TD
Intent["Human intent"] --> Router["Operating router"]
Router --> SIS["SIS: memory, standards, governance"]
Router --> ACOS["ACOS: skills, commands, agents, hooks"]
Router --> AIS["AIS: discovery, profiles, MCP, repo metadata"]
ACOS --> Modules["Creator OS / Investor OS / Business OS / Domain Intelligence"]
Modules --> Workflows["Workflows and loops"]
Workflows --> Gates["Risk, evidence, privacy, approval gates"]
Gates --> Ledger["Decision ledger and proof artifacts"]
Ledger --> SIS
```
## What This Standard Defines
| Artifact | Purpose |
|---|---|
| [STANDARD.md](./STANDARD.md) | Normative specification for compliant agentic operating systems. |
| [MODULES.md](./MODULES.md) | Module map for Creator OS, Investor OS, Business OS, Domain Intelligence, and shared substrates. |
| [AGENTS.md](./AGENTS.md) | Agent rules for this repo and a reusable doctrine for agent contracts. |
| [SKILLS.md](./SKILLS.md) | Skill format, quality bar, lifecycle, and publication rules. |
| [WORKFLOWS.md](./WORKFLOWS.md) | Canonical workflows used across modules. |
| [LOOPS.md](./LOOPS.md) | Persistent operating loops that make systems improve over time. |
| [Swarm Operating Model](./docs/SWARM_OPERATING_MODEL.md) | How to route multi-agent execution across strategy, product, risk, repo, and language lanes. |
| [README Excellence Standard](./docs/README_EXCELLENCE_STANDARD.md) | Engineering-grounded README quality bar for every module and substrate repo. |
| [Standard Deployment Runbook](./docs/STANDARD_DEPLOYMENT_RUNBOOK.md) | Safe multi-wave rollout process across clean, dirty, public, private, and substrate repos. |
| [Ecosystem Rollout Plan](./docs/ECOSYSTEM_ROLLOUT_PLAN.md) | Cross-repo rollout plan for README quality, validation, and standard artifacts. |
| [registry](./registry/agentic-operating-system-standard.json) | Machine-readable standard registry. |
| [schemas](./schemas) | JSON schemas for modules, agents, and workflows. |
| [templates](./templates) | Copyable repo, module, agent, skill, and workflow templates. |
## Compliance Levels
| Level | Name | Requirement |
|---|---|---|
| L0 | Context Pack | Has a clear README and one reusable context file. |
| L1 | Skill Pack | Adds documented skills and activation rules. |
| L2 | Workflow System | Adds workflows, loop definitions, and validation. |
| L3 | Agent OS | Adds agent contracts, safety gates, ledgers, and repo governance. |
| L4 | Standard Module | Adds schemas, public/private boundaries, release discipline, and evidence artifacts. |
| L5 | Ecosystem Node | Interoperates across SIS, ACOS, AIS, GitHub, and multiple agent runners. |
## First Modules
- [Agentic Creator OS](./modules/creator-os.md): creator production, content, media, product, and launch execution.
- [Agentic Business OS](./modules/business-os.md): founder command, revenue architecture, customer success, partner systems, and operating cadence.
- [Agentic Investor OS](./modules/investor-os.md): thesis, deal flow, diligence, memo, portfolio support, and investor communications. Regulated decisions require qualified human review.
- [Domain Intelligence System](./modules/domain-intelligence.md): Web2 and Web3 domain candidate discovery, verification, routing, and monetization workflows.
- [Agentic Asset Intelligence](./modules/asset-intelligence.md): the cross-module standard for turning digital surfaces into verified, monetizable assets.
## Repository Strategy
This repo is the standard. Product modules can live in their own repos.
| Repo Type | Purpose | Visibility |
|---|---|---|
| Standard repo | Shared language, schemas, templates, and compliance rules. | Public |
| Open module repo | Public starter implementation, examples, documentation, and community adoption. | Public |
| Pro module repo | Private plugins, customer workflows, premium agents, private evals, managed delivery. | Private |
| Customer repo | Client-specific implementation and evidence ledger. | Private |
See [Repo Strategy](./docs/REPO_STRATEGY.md).
## Use It
Validate the standard:
```bash
npm run validate
```
Assess a README against the quality bar:
```bash
python scripts/assess_readme_quality.py README.md
```
Create a new module charter:
```bash
cp templates/module-charter.md modules/my-module.md
```
Create a repo profile:
```bash
cp templates/repo-readme.md README.md
cp templates/agents.md AGENTS.md
cp templates/skills.md SKILLS.md
cp templates/readme-quality-checklist.md docs/readme-quality-checklist.md
```
## Commercial Strategy
The standard stays public. The revenue comes from modules, implementation, managed systems, audits, templates, training, and private plugin packs.
The first marketable offers should be concrete:
- Domain Portfolio Sprint.
- Creator OS setup and launch kit.
- Business OS founder command system.
- Investor OS diligence and portfolio support system.
- Agentic Operating System audit for teams.
See [Commercial Strategy](./docs/COMMERCIAL_STRATEGY.md).
## Non-Negotiables
- No hidden autonomous money movement.
- No investment, legal, medical, or tax claims without professional review.
- No public use of private customer data without approval.
- No agent action that bypasses human approval gates.
- No standard without validation, examples, and clear boundaries.
## Status
Public draft. Designed to become the shared specification behind FrankX, Starlight, Arcanea, ACOS, SIS, and specialized agentic operating system modules.