mcpbeat

Remote Compute Modal

xuzhougeng/remote-compute-modal

Explain Wisp's current Modal boundary and migrate a requested Modal workload to a supported direct SSH Run when possible. Use when an existing workflow mentions Modal, BYOC cloud compute, provider images, or cloud GPU dispatch.

369 tokens
context cost
the whole folder, loaded on every use
1
files
instructions only
0
copies elsewhere
how many repositories repackaged it
859
stars on the repo
on the repository, not the skill itself

Install

one command, takes just this skill from the repository 
npx skills add https://github.com/xuzhougeng/wisp-science --skill remote-compute-modal

The instruction itself

2 sections, as written by the author

Modal compute boundary

Wisp does not currently implement a Modal ExecutionContext or Run backend.

Only local, wsl:<distro>, and direct ssh:<alias> contexts can be passed to

run_in_context. Python receives no provider SDK, cloud credentials, image

builder, or cloud-job handle.

What to do

  • Do not submit, build, monitor, or claim to reuse a Modal image.
  • If the workload can run on a user-controlled Linux GPU host, select and

Probe an SSH context, load compute-env-setup, build the required user-space

environment there, then load remote-compute-ssh and submit a persisted Run.

  • Keep large inputs and model weights remote. Stage only small project scripts

and configuration with input_paths.

  • If no suitable SSH context exists, explain that the workload cannot be

dispatched by this Wisp build. Do not offer an untracked local SDK call as a

substitute.

A future Modal integration must add a typed execution context and a mockable

Run backend implementing resource requests, environment/image references,

keyring-backed secret binding, submit, poll, cancel, recovery, and output

harvest. That belongs in Rust, not in a Python sidecar.

How to use it

Copy the folder

Take xuzhougeng/remote-compute-modal from the repository into ~/.claude/skills for personal use, or into .claude/skills inside a project.

Check the name does not clash

The agent identifies a skill by the name field in its header. Two skills with the same name cannot sit side by side — one of them will be ignored.