A secure Python sandbox, written in Rust, for code written by AI.
Monty is a minimal Python interpreter written in Rust that runs AI-generated code in a restricted environment without direct filesystem, environment-variable, or network access. It is for AI agents that need to execute Python and type-checking tasks while allowing developers to control which host functions and objects the code can use, and the catalogue entries integrate it into agent workflows.
Latest release 1.4.0 — v1.4.0 - 2026-10-26 · 26 Oct 2026
13 files for Claude Code, Codex and OpenCode: docs-parity-reviewer, coverage, fastmod, fix-pr-comments and 9 more — 17,077 tokens loaded in every session.
AGENTS.md A 4 tok CLAUDE.md A 17,073 tok .claude/settings.json A — .agents/skills/coverage/SKILL.md A 29 tok .agents/skills/fastmod/SKILL.md A 17 tok .agents/skills/fix-pr-comments/SKILL.md A 44 tok .agents/skills/python-playground/SKILL.md A 44 tok .agents/skills/review-general/SKILL.md A 55 tok .agents/skills/review-security/SKILL.md A 57 tok .agents/skills/review-usability/SKILL.md A 52 tok .agents/skills/review-verbosity/SKILL.md A 43 tok .agents/skills/writing-style/SKILL.md A 55 tok .agents/agents/docs-parity-reviewer.md A 69 tok These files are pydantic/monty's own configuration — they tell Claude Code, Codex and OpenCode how to work on this repository, so they are not mods to install elsewhere. Copy one as a starting point and replace the parts that are about this project.