Alternatives to E2b Sandbox Mcp

Containerized environments for coding agents. Multiple agents can work independently, isolated in fresh containers and git branches. No conflicts, many experiments. Full execution history, terminal access to agent environments, git workflow. Any agent/model/infra stack.

OpenAPI-MCP: Dockerized MCP Server to allow your AI agent to access any API with existing api docs.

A Node.js MCP server that spins up isolated Docker-based sandboxes for executing JavaScript snippets with on-the-fly npm dependency installation and clean teardown

Execute any LLM-generated code in a secure and scalable sandbox environment and create your own MCP tools using JavaScript or Python, with full support for NPM and PyPI packages

Agent-first programming language: agents produce JSON AST, the compiler validates, type-checks, effect-checks, verifies contracts via Z3/SMT, and compiles to WASM. 19 MCP tools for the full compile-and-execute loop.

Connect AI agents to MATLAB — execute code, run async jobs with progress reporting, get interactive Plotly plots, expose custom .m functions as tools, and monitor via live dashboard.

MCP server for the SLURM workload manager. Submit jobs, run YAML workflows, monitor GPU resources, manage SSH profiles, and sync files to remote HPC clusters from natural language. 14 tools spanning local SLURM and SSH-remote clusters; companion CLI and FastAPI Web UI ship in the same package.

Self-hostable GPU-aware job broker for your own machines. The MCP server exposes the queue as nine tools (submit, status, logs, list, cancel, preempt, workers, job-get, worker-delete) so an agent can dispatch long-running or GPU jobs that outlive the session, route them by VRAM tier, and babysit them across sessions. pip install "jobd[mcp]".

⃣ ☁️ - Lightweight MCP server to access any API using their OpenAPI specification. Supports OAuth2 and full JSON schema parameters and request body.

Run Python code in a secure sandbox via MCP tool calls

A Javascript code execution sandbox that uses v8 to isolate code to run AI generated javascript locally without fear. Supports heap snapshotting for persistent sessions.

SWI-Prolog execution for LLMs with CLP(FD), negation-as-failure, and recursion. Benchmarked 90% vs 73% LLM-only accuracy on 30 logic problems.