fetch("https://pypi.org/simple/requests")200 OK→ isolated · scoped egress · VM destroyed/workspaceinputs/data.csvresults/output.jsonIslo runs untrusted or LLM-generated code on disposable isolated computers instead of your laptop or production network.
Execute LLM-generated scripts on throwaway isolated computers with egress policies, then discard the machine when the job is done.
Give your coding agent the Islo skill. It builds and runs autonomous agents on Islo for you.
fetch("https://pypi.org/simple/requests")200 OK→ isolated · scoped egress · VM destroyed/workspaceinputs/data.csvresults/output.jsonPay only while the computer runs: CPU + memory + disk. No seat licenses, no idle fees.
From our blog
Generated scripts may be wrong or malicious. Islo runs them on disposable computers with scoped egress instead of your laptop or production network.
Yes. Pipe generated code into an isolated computer through the CLI or SDK and collect stdout, stderr, and artifacts.
You can stop and remove the computer, or snapshot state if you need to inspect results later.
Only through the egress rules you configure. Gateway policies can allow required hosts and block everything else without placing credentials inside the computer.
Yes. Collect stdout, stderr, and artifacts through the CLI or SDK, and keep or snapshot the computer when you need a longer investigation.
$ uv add islo openaifrom openai import OpenAI
from islo import Islo
llm = OpenAI()
code = llm.responses.create(
model="gpt-4.1",
input="Write Python to validate this uploaded CSV.",
).output_text
islo = Islo()
sandbox = islo.sandboxes.create_sandbox(name="llm-output")
islo.sandboxes.exec_in_sandbox(
sandbox_name=sandbox.name,
command=["python", "-c", code],
)