"Fresh" matching MCP servers:

Ephemeral MicroVM-isolated code execution for AI agents. Run Python, Node, or bash — fresh hardware-isolated VM per call, hard-purged after. No state persists between calls.

Enables users to create, inspect, analyze, format, and chart Excel workbooks by having the client LLM write Python that runs in a fresh, network-restricted Docker container. Exposes a single execute_python tool with access to openpyxl, pandas, numpy, and matplotlib, persisting files between calls.

Enables executing submitted code snippets in a fresh isolated V8 isolate, capturing stdout, stderr, return values, and failures under enforced resource limits.

Enables agents to perform recorded, snapshot-evidenced computer-use and browser-use in a fresh, disposable VM, with interruptive actions judged by the agent.

Enables AI coding agents to work inside a disposable Windows guest desktop, where they can run shell commands, launch applications, inspect and operate native GUI controls, perform browser DOM actions, and transfer files. Fresh observations, PNG screenshots, and hashed artifacts let agents verify results without touching the host desktop.

Exposes a single execute tool so agents can discover, inspect schemas for, and compose calls to connected upstream MCP servers by writing JavaScript that runs in a fresh sandboxed QuickJS VM. Results from combined tool calls are returned as text or JSON, keeping large tool catalogs out of the agent's context.

Enables AI agents to run JavaScript in fresh V8 isolates and shell commands under macOS seatbelt sandboxing, with per-call permissions for network, CPU, memory, disk, and host mounts. It also supports persistent writable copies with explicit sync or discard for safe host directory edits.

stateless MCP server — fresh instance per request, infinite horizontal scaling