AI-CEO Framework skill
A Claude Code template: 15 department agents, five skills and an approval queue for outward actions such as emails and invoices.
- 61 ★ on the repository
- Last commit
- MIT
Broad access
Why this level
- Agents handle finance, contracts, emails and deployments, some jobs run on a schedule, and protection rests on the approval queue and a permissions file.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/ai-ceo-frameworkDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- Agents handle finance, contracts, emails and deployments, some jobs run on a schedule, and protection rests on the approval queue and a permissions file.
This is third-party code. Review the repository files before installing.
What it does
The framework turns a Claude Code project into a one-person company: a `CLAUDE.md` file parses plain requests and routes them to department agents. `agents/` holds 15 agents: CTO, marketing, finance, sales, legal, support, HR, publishing, growth, consulting, business development, tax, a morning digest and a setup wizard. `skills/` has five skills: six-gate hypothesis validation, blog writing and polishing, automation upgrades and cover generation. Any outward action (emails, posts, invoices, deployments) first lands as a draft in `approval-queue.md` and runs only after the `/ai-ceo:approve` command. The `setup.sh` wizard asks about the company and budget.
Who it is for. Founders running a business alone who want to hand department routine to Claude Code agents while keeping approval for themselves.
Good fit when
- A morning digest across all departments is needed: the morning digest agent handles it.
- Before launching a new product you want to run six-gate hypothesis validation.
- Client invoices and emails must go out only after your approval in the queue.
- A contract review or an open source license audit is needed from the legal agent.
Not a fit when
- A ready product without setup is needed: the framework needs answers to questions and your own tools connected.
- The company is large and roles already belong to people.
- Unattended automation is unacceptable: part of the jobs run on a schedule through `launchd`.
Example request
What is our status today? Prepare a morning digest across departments and put the client email into the approval queue.Limitations
The framework targets Claude Code: the Max plan and metered usage for headless runs affect cost, and that part grows with automation. The README gives verifiable figures as of November 11, 2026: in production since February 16, 2026, 90 percent of commits made by automation, 32 scheduled jobs. The author notes that earlier figures were overstated. The real working copy has 17 agents and 13 skills, while 15 and 5 are published here. Action rights are set in `steering/permissions.md`.
How to disable. Remove the framework `.claude/` directory and the `CLAUDE.md` file from the project, and stop and delete the `launchd` jobs created by `setup.sh` by hand.
FAQ
How do I install the framework?
Copy `.claude/` and `CLAUDE.md` into your project with `cp -r`, then run `bash .claude/setup.sh` and answer the company questions.
What does the setup wizard ask?
Company name, your name, business description, products, tech stack, external tools, priority departments and AI budget.
How does the approval queue work?
An agent puts a draft in `approval-queue.md`. You approve it with `/ai-ceo:approve <id>` or reject it with `/ai-ceo:reject <id> "reason"`, and only approved items execute.
What does hypothesis validation do?
It runs six gates: idea origin, market existence, interviews, fact scoring, willingness to pay and a minimum test.
Which commands exist?
For example `/ai-ceo:morning` for the morning digest, `/ai-ceo:status` for status and `/ai-ceo:dev:sprint` for a development sprint.
Discussion
No questions or reviews yet.
Related
SkillsA template for Claude Code projects: parallel agent swarms, workflow skills, safety hooks and rules in one install.
AI Builder Club Skills
AI Builder Club — Skills
A Claude Code plugin marketplace to make a repo agent-ready and run compounding agent loops
A personal ~/.claude for Claude Code: twelve subagents, review commands, worktree and goal skills and a Go status line.
A Claude Code skill that takes a founder from idea to a live Claude Managed Agent: interview, v0 build, launch, grading and scheduling
Multi-agent orchestration for Claude Code: teams, roles, autopilot and task routing across skills and subagents
A Claude Code workspace folder: hooks with a local Ollama detect the project type, pick skills and remind you to update the docs
README in short
The README opens with measurements from a running company and corrects earlier overstated figures. It then covers the pack contents, an architecture with department routing and the quick start order.