Babysitter
An orchestrator for agents: the workflow is defined in code, every step is enforced, key points need human approval and decisions go into an immutable journal
High risk
We rate an entry high when the tool writes to external systems, handles money, production databases or secrets, or runs arbitrary commands. The CLI installs it only with your consent.
Why this level
- Directs agents and runs steps that execute code
- Can run complex multi-step processes autonomously between checkpoints
Install
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/babysitterDetects the agents on your machine, checks the risk and pins the version.
Other ways to install
Install the tool
npm install -g @a5c-ai/babysitterInstall the CLI: npm install -g @a5c-ai/babysitter. For Claude Code add the plugin: claude plugin marketplace add a5c-ai/babysitter-claude, then claude plugin install --scope user babysitter@a5c.ai.
Other ways from the author
npm install -g @a5c-ai/babysitterThe global orchestrator CLI.
This is third-party code. Review the repository files before installing.
What it does
Babysitter makes teams of agents follow a defined process strictly. The workflow is described in code, and the orchestrator ensures each step runs and quality gates pass before moving on. At checkpoints it requires human approval, and it records every decision in an immutable journal. It installs as a CLI via npm, with plugins for Claude Code, Codex and Cursor for specific agents.
Who it is for. For teams and developers running complex multi-step tasks on agents who want predictability.
Good fit when
- You need a strict, repeatable process for agents
- You need quality gates and human approval between steps
- You need a journal of every decision
Not a fit when
- The task is simple and needs no orchestration
- You want agent freedom without strict rails
Example request
Run this task through Babysitter step by step with quality gates and a stop for my approvalLimitations
The process must be described in code, which takes setup time. The orchestrator runs steps that execute code, so use it in a controlled environment. The agents themselves call their models, whose availability from Russia should be checked separately.
How to disable. Remove the @a5c-ai/babysitter package via npm and uninstall the added plugins from your agents.
Security check
- Directs agents and runs steps that execute code
- Can run complex multi-step processes autonomously between checkpoints
README in short
The README describes Babysitter as a deterministic orchestrator for teams of agents, with the process defined in code. It installs via npm as a global CLI, with separate plugins for Claude Code, Codex and Cursor. It lists quality gates, checkpoints with human approval and an immutable decision journal. It also mentions a runtime package for orchestration and build commands.
FAQ
How is the process defined?
In code: you set the steps and checks, and the orchestrator enforces them and keeps a journal.
Does it replace the agent?
No, it directs your agents and hosts rather than replacing them.
Related
Open-source personal AI assistant on your own machine: answers in Telegram, Slack, Discord and WhatsApp, extended with skills and plugins
A self-improving agent from Nous Research with a TUI, messaging gateway, cron jobs and skills it writes itself
An open source coding agent for the terminal and desktop with build and plan modes
Open prompt library with a Claude Code plugin, MCP server and CLI: search, fetch and improve prompts and skills from an agent