swe-rex
Sandboxed code execution for AI agents, locally or on the cloud.
Decision gist · record as of 2026-08-14
Yes, if you are building AI agents that need to execute shell commands in sandboxed or remote environments. The package solves a real infrastructure abstraction problem and is actively maintained with no known vulnerabilities. Install friction is low and the permissive MIT license removes legal friction. Start with the base install; add optional dependencies only if you need Modal, Fargate, or Daytona support.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires Python >=3.10.
- For cloud platforms (Modal, Fargate, Daytona), additional optional dependencies and platform credentials are needed.
- Low friction install with a pure-Python wheel.
License · maintenance · safety
permissive license (permissive) — MIT license (permissive) allows unrestricted use, modification, and distribution with minimal legal friction for commercial or private projects.
last release 2025-08-14 (365 days) · last repo commit 2026-08-10 · 570 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 3,151,672 downloads/mo, #2,724 on PyPI
Alternatives
Verify before relying
pip install swe-rex
from swe_rex import LocalExecutor
executor = LocalExecutor()
result = executor.run_command('echo hello')- Exact API surface and how to instantiate executors for different platforms (Docker, AWS, Modal) beyond the install extras shown.
- Whether interactive tools like ipython or gdb require special configuration or just work out of the box.
- Performance characteristics and latency overhead when running commands remotely vs. locally.
- Limits on parallel agent count or resource consumption when scaling to many concurrent sessions.
What it is and what it does
SWE-ReX is a runtime abstraction layer that lets AI agents execute shell commands in sandboxed environments without caring whether they run locally, in Docker, on AWS, or on Modal. It handles the messy details of spawning sessions, detecting command completion, capturing output and exit codes, and managing multiple parallel shells—so your agent code stays the same whether you're testing locally or scaling to dozens of parallel runs.
The package is built on fastapi and uvicorn, suggesting it exposes a web service interface for remote execution. It integrates with pexpect and bashlex to parse and interact with shell sessions, and uses pydantic for configuration validation. It's designed for developers building AI agents that need to run arbitrary commands in controlled environments without reimplementing infrastructure plumbing.
Use it for
- Let an AI agent run code-fixing commands in an isolated Docker container without hardcoding container-specific logic in your agent.
- Evaluate an AI agent on a benchmark by spinning up parallel shell sessions and running the agent on each without managing infrastructure.
- Build a multi-tool agent that uses ipython, gdb, and bash interactively, with SWE-ReX handling session multiplexing and output parsing.
- Deploy an agent to Modal or AWS and have it execute shell commands remotely while your agent code remains unchanged.
- Run many agents in parallel locally or on the cloud without writing custom process management or container orchestration code.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you are building AI agents that need to execute shell commands in sandboxed or remote environments.
The package solves a real infrastructure abstraction problem and is actively maintained with no known vulnerabilities. Install friction is low and the permissive MIT license removes legal friction. Start with the base install; add optional dependencies only if you need Modal, Fargate, or Daytona support.
Install
swe-rex on PyPI
Before you install
Low friction install with a pure-Python wheel. Active maintenance (last commit 2026-08-10) and recent releases (1.4.0 on 2025-08-14) suggest ongoing support. Eight runtime dependencies including fastapi, uvicorn, and pydantic indicate a web-service-oriented architecture.
Requires Python >=3.10. For cloud platforms (Modal, Fargate, Daytona), additional optional dependencies and platform credentials are needed.
License in practice
MIT license (permissive) allows unrestricted use, modification, and distribution with minimal legal friction for commercial or private projects.
Quickstart
pip install swe-rex
from swe_rex import LocalExecutor
executor = LocalExecutor()
result = executor.run_command('echo hello')Verify before relying
- Exact API surface and how to instantiate executors for different platforms (Docker, AWS, Modal) beyond the install extras shown.
- Whether interactive tools like ipython or gdb require special configuration or just work out of the box.
- Performance characteristics and latency overhead when running commands remotely vs. locally.
- Limits on parallel agent count or resource consumption when scaling to many concurrent sessions.
Package facts
| License | permissive license permissive |
| Python support | Supports the current Python release >=3.10 |
| Install friction | Low. Pure-Python wheel |
| Runtime dependencies | 8 packagesfastapiuvicornrequestspydanticpexpectbashlexpython-multipartrich |
| Maintenance | Actively maintained 365 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 3,151,672 / month, #2,724 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Intended Audience :: DevelopersLicense :: OSI Approved :: MIT LicenseOperating System :: OS IndependentProgramming Language :: Python :: 3 :: OnlyProgramming Language :: Python :: 3.11 |
Evidence: swe_rex-1.4.0-py3-none-any.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “ai agent shell execution”
- swe-rexSWE-ReX provides a runtime interface for AI agents to execute shell…
- mistral-vibeMistral Vibe is a command-line coding assistant that lets you…
- strands-agents-builderStrands Agent Builder is a terminal-based toolkit for building,…
Give your agent the search over MCP, or paste the wish link into any chat.
More Artificial Intelligence packages
LiteLLM provides a unified Python interface to call 100+ LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, and others) using OpenAI-compatible API format, available as both a Python SDK and a self-hosted AI Gateway proxy server.
Install it if you need to work with multiple LLM providers or want to centralize LLM routing in your organization.
Client library and CLI tool for downloading, uploading, and managing models, datasets, and repositories on the Hugging Face Hub platform.
Install it if you work with Hugging Face Hub models or datasets.
LangChain provides a framework for building agents and LLM-powered applications by composing language models, tools, and memory through a unified API that abstracts over multiple model providers.
hf-xet provides chunk-based deduplication and efficient file transfer for the Hugging Face Hub, enabling faster uploads and downloads of large files with local disk caching.
Tokenizers converts raw text into token sequences for NLP models, with support for training custom vocabularies and using pre-built tokenizers (BPE, WordPiece) optimized for speed via Rust.
Transformers provides a unified framework for loading, fine-tuning, and running state-of-the-art pretrained models across text, vision, audio, video, and multimodal tasks using PyTorch, JAX, or TensorFlow.
Install it if you need to run or train any transformer-based model for NLP, vision, audio, or multimodal tasks.
See also mini-swe-agent · swesmith · pykaos · swebench · spotinst-agent · spotinst-agent-2 · e2b · pydantic-ai-backend · harbor · google-agents-cli