Skip to main content

DSPy CodeInterpreter implementation using Monty, a secure Python interpreter written in Rust

Project description

dspy-monty-interpreter

DSPy CodeInterpreter implementation using Monty, a secure Python interpreter written in Rust.

The Monty team points out, "This project is still in development, and not ready for the prime time." It uses a small subset of the standard library (sys, os, typing, asyncio, re, datetime, json) and can't yet define classes or use match statements.

That said: Monty is fast. For many RLM use cases, Monty is my daily driver.

Installation

pip install dspy-monty-interpreter

Requires pydantic-monty>=0.0.10.

Usage

import dspy
from dspy_monty_interpreter import MontyInterpreter

interpreter = MontyInterpreter()
rlm = dspy.RLM("context -> answer", interpreter=interpreter)
result = rlm(context="What is 2 + 2?")

Standalone usage

from dspy_monty_interpreter import MontyInterpreter

interp = MontyInterpreter()

# Basic execution
interp.execute("x = 42")
interp.execute("print(x + 8)")  # returns "50"

# State persists across calls
interp.execute("def double(n):\n    return n * 2")
interp.execute("double(21)")  # returns "42"

# With tools
def lookup(key: str) -> str:
    return "some value"

interp = MontyInterpreter(tools={"lookup": lookup})
interp.execute('result = lookup(key="foo")\nprint(result)')

Filesystem mounts

Filesystem access is disabled by default. Use mounts= to expose a host directory to sandboxed code under a virtual path:

from dspy_monty_interpreter import MontyInterpreter, MountDirectory

# Read-only mount: sandbox can read but not write
interp = MontyInterpreter(
    mounts=MountDirectory("/data", "/host/path/to/data", mode="read-only")
)
interp.execute("from pathlib import Path\nPath('/data/input.csv').read_text()")

# Overlay mount: reads fall through to host; writes stay in memory
interp = MontyInterpreter(
    mounts=MountDirectory("/work", "/host/project", mode="overlay")
)

Modes: "read-only", "read-write", "overlay". Pass a list of MountDirectory for multiple mount points.

Why Monty?

  • Fast: Microsecond startup (no subprocess, no WASM bootstrap)
  • Secure: No network or environment access; filesystem access is opt-in via mounts
  • Lightweight: Pure Rust, no Deno/Pyodide dependency

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

dspy_monty_interpreter-0.2.0.tar.gz (13.2 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

dspy_monty_interpreter-0.2.0-py3-none-any.whl (5.7 kB view details)

Uploaded Python 3

File details

Details for the file dspy_monty_interpreter-0.2.0.tar.gz.

File metadata

  • Download URL: dspy_monty_interpreter-0.2.0.tar.gz
  • Upload date:
  • Size: 13.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.4

File hashes

Hashes for dspy_monty_interpreter-0.2.0.tar.gz
Algorithm Hash digest
SHA256 97657536a3b4f069b6f1a32f02545191147bb68d4c339cf03dd1e781e0a25f5a
MD5 5f3c22bd8ac47eb50ab51fc5e764f0a8
BLAKE2b-256 b3fc6556e5183fe231b5b96e4798302f0ed0c12414db1e6736c5c164d7b7e923

See more details on using hashes here.

File details

Details for the file dspy_monty_interpreter-0.2.0-py3-none-any.whl.

File metadata

File hashes

Hashes for dspy_monty_interpreter-0.2.0-py3-none-any.whl
Algorithm Hash digest
SHA256 2a2fe2b13ca6b8b31e3514d43e4633dbf46beea48d1f9f2a0682a4c21b65b236
MD5 9f81d29e363bf0ddac8223112135a8fe
BLAKE2b-256 0dde14e9e433734d9db76dd2dc040cc34799c2662e90662bd3904aefc9c828b9

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page