The open standard for AI agent organizations — define, validate, and deploy AI teams with ranked roles, communication protocols, and escalation rules.
No framework lock-in. No library required. Just markdown — implementable in any language, on any platform, with any LLM.
# 1. Install the CLI
pip install corpai
# 2. Clone the spec
git clone https://github.com/Arigitshub/CorpAI && cd CorpAI
# 3. Validate and explore
corpai lint # validate all 46 roles
corpai graph # visualize the org tree
corpai info # org summarygraph TD
OWNER["OWNER — Human Principal"]
CEO["[L5] CEO"]
CFO["[L5] CFO"]
CTO["[L5] CTO"]
COO["[L5] COO"]
CMO["[L5] CMO"]
OWNER --> CEO
CEO --> CFO & CTO & COO & CMO
CFO --> Finance["Finance\nDirector → Analyst → Auditor → Tracker"]
CTO --> Engineering["Engineering\nDirector → Lead → Engineer → QA"]
CTO --> Security["Security\nDirector → Manager → Analyst → Monitor"]
CTO --> DataAI["Data/AI\nDirector → ML Lead → Scientist → Engineer"]
COO --> Operations["Operations\nDirector → PM → Analyst → Coordinator"]
COO --> HR["HR\nDirector → Manager → Specialist → Onboarding"]
COO --> CS["Customer Success\nDirector → Manager → Specialist → Support"]
CMO --> Marketing["Marketing\nDirector → Content → Writer → Growth"]
CEO --> Legal["Legal\nDirector → Compliance → Contract → Policy"]
Full charts → spec/diagrams/org-chart.md
- Language-agnostic — the spec is markdown. Implement it in Python, TypeScript, Go, or whatever you ship in.
- Production-ready structure — 46 defined roles across 10 departments, with ranks (L1–L5), escalation rules, and cross-department protocols baked in.
- CLI-validated —
corpai lintcatches broken chains, missing fields, and spec violations before you deploy.
| Repo | What it is |
|---|---|
| CorpAI | The spec — this repo |
| corpai-cli | CLI to lint, visualize, simulate (pip install corpai) |
| corpai-runtime | Provider-agnostic agent execution engine |
| corpai-portal | Interactive org chart dashboard |
| File | What it defines |
|---|---|
| spec/ranks.md | L1–L5 rank system |
| spec/communication.md | Message types, format, priority levels |
| spec/escalation.md | Escalation triggers and OWNER notifications |
| spec/lifecycle.md | Agent lifecycle: Proposed → Active → Decommissioned |
| spec/cross-department.md | How departments interact |
| spec/multi-org.md | Federation protocol for inter-org communication |
| spec/message-examples.md | Real message examples |
| spec/glossary.md | Every term, defined |
| Department | Roles |
|---|---|
| Executive | OWNER, CEO, CFO, CTO, COO, CMO |
| Engineering | Director, Team Lead, Senior Engineer, Engineer, QA Lead, QA Tester |
| Finance | Director, Financial Analyst, Auditor, Budget Tracker |
| Marketing | Director, Content Lead, Growth Specialist, Brand Strategist, Content Writer |
| Operations | Director, Project Manager, Process Analyst, Coordinator |
| Legal | Director, Compliance Specialist, Contract Reviewer, Policy Checker |
| HR | Director, HR Manager, Talent Specialist, Onboarding Agent |
| Security | Director, Security Manager, Threat Analyst, Monitor Agent |
| Data/AI | Director, ML Lead, Data Scientist, Data Engineer, Data Processor |
| Customer Success | Director, CS Manager, Account Specialist, Support Agent |
Add the badge to your repo → BADGE.md
Apply for CorpAI Certified status → CERTIFIED.md
PRs and proposals welcome. Read CONTRIBUTING.md to add roles, propose spec changes, or submit a certified implementation.
GitHub Discussions — questions, ideas, proposals