LibreAI / Dev Tools / pydantic-deepagents
// devtools Β· vstorm-co/pydantic-deepagents

pydantic-deepagents

Self-hosted terminal AI assistant and Python agent framework on Pydantic AI.

πŸ’Ž Hidden gemβ˜… 1.1k starsβ‘‚ 134 forksMITupdated 2026-09
β–¦ visual β€” this project's README ships no embeddable screenshot,
so here's a placeholder. The guide below is complete regardless.
// project facts

Project facts

pydantic-deepagents is an open-source dev tools project by vstorm-co (vstorm-co/pydantic-deepagents on GitHub). It has 1,076 GitHub stars and 134 forks, is released under the MIT license, and was last updated on GitHub on 15 September 2026.

GitHub stars
1,076
Forks
134
License
MIT
Last updated on GitHub
15 September 2026
Category
Dev Tools
Listing
Hidden gem β€” a shorter page for a lesser-known project
Official site
Official docs
// how it works

How it works

  1. Clone the repo

    Get the source and move into the project folder.

  2. Add your model key

    Point it at the LLM provider of your choice.

  3. Launch the assistant

    Start the terminal agent inside your project directory.

  4. Build on the framework

    Use the Python API to script multi-agent workflows of your own.

// features

Why people love it

Terminal assistant

A Claude Code-style agent that lives in your shell.

Tool calling

Reads, edits, and runs code through real tools.

Sandboxed execution

Docker sandbox keeps agent-run commands contained.

Multi-agent teams

Subagents split work across parallel tasks.

MCP support

Plugs into Model Context Protocol servers and tools.

Skills and checkpoints

Reusable skills plus restorable session checkpoints.

Advertisement
// run it

Run it yourself

  1. Step 1 β€” Clone the repo

    Download the source and enter the folder.

    git clone https://github.com/vstorm-co/pydantic-deepagents.git
    cd pydantic-deepagents
  2. Step 2 β€” Install it

    Install the Python package and its dependencies.

    pip install -e .
  3. Step 3 β€” Launch

    Start the assistant; see the README for your model config.

    # set your model API key first, then follow the README launch command
// use cases

What you can build

Advertisement
// faq

Questions, answered

What is pydantic-deepagents?

An open-source, self-hosted terminal AI assistant plus the Python framework behind it, built on Pydantic AI at vstorm-co/pydantic-deepagents.

Does it require a specific model?

No. It runs on Pydantic AI, so you can point it at any supported model provider.

Is it free for commercial use?

Yes. It is MIT licensed.

How is it different from Claude Code?

It is self-hosted and open source, so your code and prompts stay on your own machine.

What do I need to run it?

Python, a model API key, and Docker if you want sandboxed execution.

// keep exploring

Similar tools

Local LLM

Skills-Manager

Write an agent skill once, sync it to 32 AI coding tools.

β˜… 1.0k starsMIT
View guide β†’
Agents

claude-code-from-scratch

Rebuild Claude Code’s core in ~5,000 lines: an 11-chapter coding-agent masterclass.

β˜… 2.7k starsMIT
View guide β†’
Dev Tools

claude-code-security-review

An AI-powered security review GitHub Action using Claude to analyze code changes for security vulnerabilities.

β˜… 6.3k starsMIT
View guide β†’
Dev Tools

sandboxd

Self-hosted AI app builder β€” an agent builds apps in isolated sandboxes.

β˜… 961 starsMIT
View guide β†’
Dev Tools

Tabby

Self-hosted AI coding assistant

β˜… 33.9k starsCustom (check repo)
View guide β†’
Dev Tools

git-iris

Self-hosted dev tools: an intelligent agent that understands your code and cra.

β˜… 41 starsApache-2.0
View guide β†’
// alternatives

Alternatives in Dev Tools

The most-starred other Dev Tools projects in the LibreAI catalogue β€” the realistic shortlist if pydantic-deepagents is not the right fit.

Get pydantic-deepagents

Free and open β€” grab the source, star the repo, and start creating.