modal-research-compute

name	modal-research-compute
description	Use when a research or engineering task needs automatic heavy-compute routing through the local broker for Modal-backed remote CPU, high-memory CPU, or GPU execution.
user-invocable	true
disable-model-invocation	true
metadata	{"openclaw":{"requires":{"bins":["python3","modal"]}}}

Modal Research Compute

Use this skill when the task is about:

• exhaustive search
• object enumeration
• counterexample hunting
• large parameter sweeps
• remote execution of generated experiment code
• GPU-suitable document, embedding, reranking, or tensor workloads

This skill is the OpenClaw integration layer for the local research_compute broker.

When to prefer this skill

• the user wants Modal involved automatically
• the local machine is CPU, memory, disk, or GPU constrained for the requested workload
• the workload is long-running enough that remote execution is a better fit

Core workflow

1. If local resources matter, run get-available-resources.
2. Build a broker manifest JSON for the task.
3. Run broker plan.
4. If the plan stays within policy, run broker submit.
5. Use wait and fetch to retrieve results and logs back to local storage.

Runtime commands

Doctor:

exec: /workspace/skills/modal-research-compute/run_modal_research_compute.sh doctor

Plan:

exec: /workspace/skills/modal-research-compute/run_modal_research_compute.sh plan /path/to/job.json

Submit:

exec: /workspace/skills/modal-research-compute/run_modal_research_compute.sh submit /path/to/job.json

Wait:

exec: /workspace/skills/modal-research-compute/run_modal_research_compute.sh wait <job_id>

Fetch:

exec: /workspace/skills/modal-research-compute/run_modal_research_compute.sh fetch <job_id> --dest /path/to/output

Deploy:

exec: /workspace/skills/modal-research-compute/run_modal_research_compute.sh deploy

Operational notes

• The broker is the decision boundary. Do not call Modal directly from the normal OpenClaw flow when the broker can handle the task.
• CPU-heavy combinatorial workloads should default to remote CPU or high-memory CPU, not GPU.
• GPU use should be explicit in the manifest or clearly justified by the workload.
• doctor and plan work without a deployed Modal app. submit, wait, fetch, and deploy need the host to be Modal-ready.
• Linux hosts become Modal-ready after python3 -m pip install --user --upgrade modal and modal token set or modal token new.
• Broker state persists under the OpenClaw research data tree, while fetched outputs materialize under the caller workspace by default.