agentleFS
phnx-labsphnx-labs/rush-mcp/skills/run-in-parallel/SKILL.md

run-in-parallel

Skill0 starsChanged 4 days agoMITView on GitHub

Run many Rush tasks at once, one cloud computer each, and gather the results. Use when someone has a list of independent pieces of work (issues to fix, repositories to update, files or pages to research, flaky tests to chase), or wants several approaches to one problem tried side by side.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Start them: Call create_session once per item, each with its own fresh UUID, written yourself, as request_id, and the same repo, project or …

    line 18
  2. Very specific:

    Gather: Call wait_for_sessions with all the ids (up to 20 per call), until: "any" and timeout_seconds: 50.

    line 23
  3. Very specific:

    Several approaches to one problem: Start one session per approach with the same task and the approach named in it, then compare the results …

    line 31

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Run tasks in parallel with Rush

Each Rush session has its own computer and shares nothing with the others, so independent work runs side by side instead of in a queue.

  1. Split the work. Make one task per item that can be done without the others' results. Each task is self-contained: the goal, its inputs, and the same definition of done for every item ("open one pull request and do not merge it", "write findings.md"; leave out "do not merge" only when the person asked for merges). Work that touches the same files belongs in one task, not several racing pull requests. Show the person the list and confirm before starting more than five.
  2. Start them. Call create_session once per item, each with its own fresh UUID, written yourself, as request_id, and the same repo, project or agent where they share one. Keep a table of item to session id. Sessions past the account's runs-at-once limit start queued and begin as others finish.
  3. Gather. Call wait_for_sessions with all the ids (up to 20 per call), until: "any" and timeout_seconds: 50. Report each session as it settles, then call again with the ones still running.
  4. Summarize. One row per item: its status, a one-line answer, and the pull request or file link. List failed and needs_review items first, with their reasons. To redo one, send a follow-up to that session with send_session_message rather than starting over.

Several approaches to one problem. Start one session per approach with the same task and the approach named in it, then compare the results side by side: what each changed (get_session_changes), whether tests pass, and the trade-off. Let the person pick; offer cancel_session for the others only when they ask.

Refusals: see hand-off-a-task under "When Rush refuses".

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.