pydantic-deepagents
pydantic-deepagents (pydantic-deep)
An open-source terminal AI assistant and Python framework with a Docker sandbox, agent tools and subagents
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
- Runs code and commands, even if in a Docker sandbox
- Sends data to external models with keys
Install
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/pydantic-deepagents-pydantic-deepDetects the agents on your machine, checks the risk and pins the version.
Other ways to install
Install the tool
pipx install pydantic-deepWithout pipx, pip install pydantic-deep works too.
Install the assistant with the script curl -fsSL https://raw.githubusercontent.com/vstorm-co/pydantic-deep/main/install.sh | bash or via pip install pydantic-deep. For the CLI elsewhere use pip install "pydantic-deep[cli]".
Other ways from the author
curl -fsSL https://raw.githubusercontent.com/vstorm-co/pydantic-deep/main/install.sh | bashOn Windows or manually: pip install "pydantic-deep[cli]".
This is third-party code. Review the repository files before installing.
What it does
pydantic-deepagents is a terminal AI assistant and, at the same time, the Python framework it is built on. It can call tools, run code in a Docker sandbox, assemble subagent teams, keep checkpoints and connect MCP. There is a headless mode for running without a UI and browser automation via Playwright. It installs with a script or via pip, runs as a TUI and headless. The project positions itself as a self-hosted open-source counterpart to Claude Code.
Who it is for. For developers who want an open-source self-hosted coding assistant with a sandbox and subagents.
Good fit when
- You want an open-source terminal assistant with sandboxed code execution
- You need subagent teams and checkpoints
- You need a headless mode for automation
Not a fit when
- You already use another coding assistant and do not want to switch
- You have no access to the required model APIs
Example request
Run the task in the Docker sandbox and split it across subagentsLimitations
It needs model API keys, and some providers are not reachable from Russia without VPN. The sandbox relies on Docker, and the assistant runs code and commands, so access should be controlled. Installation requires Python.
How to disable. Remove the package with pip uninstall pydantic-deep.
Security check
- Runs code and commands, even if in a Docker sandbox
- Sends data to external models with keys
README in short
The README presents pydantic-deep as a self-hosted open-source counterpart to Claude Code: a terminal assistant and Python framework with tool calling, a Docker sandbox, subagent teams, checkpoints and MCP. Installation is via a script or pip, with a headless mode and browser automation through Playwright. It covers TUI and headless modes and the sandbox's named workspaces. MIT licensed.
FAQ
Is it a framework or a ready assistant?
Both: a ready terminal assistant and the Python framework it is built on.
How is code execution isolated?
Through a Docker sandbox with named workspaces where installed packages persist.
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