agentleFS
VercelKnown publishervercel/turborepo/skills/add/SKILL.md

add

Skill31k starsChanged 33 days agoMITView on GitHub

Add Turborepo to an existing repository in any toolchain Turborepo supports. Use when asked to add, introduce, or set up Turborepo or turbo in a repository that does not use it yet.

10 other projects use a version of this.

Trending

Quiet since July.about the whole repository

Dec 2025: 1 mentionJan 2026: 3 mentionsFeb 2026: 2 mentionsMar 2026: 2 mentionsApr 2026: 1 mentionMay 2026: 5 mentionsJun 2026: 17 mentionsJul 2026: 13 mentionsAug 2026: 0 mentionsSep 2026: 0 mentionsOct 2026: 0 mentionsNov 2026: 0 mentions
Dec 2025Jun 2026Nov 2026
Mentions a month of the vercel/turborepo repository that our weekly crawl found. There is no adoption data for this file yet.

Most recently taken up by

Who's talking about vercel/turborepo

The biggest 5 of 88 mentions linking to the repository that our crawl found. Each line is their own headline.

See all 88

Talked about on GitHub, Bluesky, Hacker News, Stack Overflow, the web and YouTube (88 times).

On GitHub · 50

45 more on GitHub

On Bluesky · 30

25 more on Bluesky

On the web · 2

On YouTube · 1

Used with

Projects with it also use note.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • MicrosoftSkill

    note

    Use this skill to record progress, partial findings, hypotheses, and page-anchored evidence while investigating a long document. Call Note aggressively — any time you read a page and confirm, narrow, or update your understanding of the user's question. Triggers include "jot that down", "save this finding", "note that page 7 says X", or (more often) implicit internal signals — you just learned something and want it recoverable later. Notes persist in the session and Review can select among them. Never paraphrase notes back from memory; write them as you go.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

12 rules. 0 are common, 6 are distinctive, and 6 are very specific to this project.

  • 0

    Common

    Most of the 9 edited versions we read keep these.

  • 6

    Distinctive

    This author's own choices. Worth a look.

  • 6

    Very specific

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

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

    Add Turborepo to an existing repository for selected tasks, without changing what existing commands do.

    line 8
  2. Distinctive:

    Read these before changing files, and treat them as the source of truth for configuration syntax and toolchain prerequisites:

    line 12
  3. Distinctive:

    Stop and explain why, leaving the repository unchanged, if:

    line 24
  4. Distinctive:

    Read the manifests, workspace definition, lockfiles, root scripts or task-runner targets, package dependency relationships, and CI …

    line 36
  5. Very specific:

    Keep existing root entry points.

    line 51
  6. Distinctive:

    Add any prerequisites from the toolchain guide that are missing, such as a workspace name or toolchain declaration.

    line 57
  7. Very specific:

    Create a root turbo.json for the selected tasks only:

    line 59
  8. Very specific:

    Add dependsOn, inputs, outputs, and environment variables only when the repository shows they are needed.

    line 62
  9. Very specific:

    If you add custom inputs, include "$TURBO_DEFAULT$".

    line 63
  10. Very specific:

    Add the dependency directory and .turbo to .gitignore if they are missing.

    line 68
  11. Distinctive:

    Only the intended packages and tasks are selected, and no package that defines a task is missing.

    line 74
  12. Distinctive:

    Commit, push, or open a pull request only if the user asks.

    line 90

“Common” is measured: kept by at least 6 in 10 of the 9 edited versions of this file we could read. “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

Add Turborepo

Add Turborepo to an existing repository for selected tasks, without changing what existing commands do. In workspace repositories, configure workspace package tasks. In single-package repositories, configure root tasks.

Documentation

Read these before changing files, and treat them as the source of truth for configuration syntax and toolchain prerequisites:

Once turbo is installed, prefer its bundled docs/README.md, which matches the installed version. If you cannot verify the docs, stop instead of relying on remembered syntax.

When to stop

Stop and explain why, leaving the repository unchanged, if:

  • the Git checkout has uncommitted changes;
  • Turborepo cannot discover packages from the repository's existing manifests;
  • adding Turborepo would require a new toolchain, package manager, or manifest type, or would require moving packages;
  • one toolchain has conflicting lockfiles;
  • you cannot tell from the repository and CI how any selected task is ordered, what it outputs, or what environment it needs.

Workflow

1. Analyze

Read the manifests, workspace definition, lockfiles, root scripts or task-runner targets, package dependency relationships, and CI configuration. Summarize the toolchains, layout, candidate tasks, observed dependencies, outputs, environment variables, and anything you could not infer.

A direct request to add Turborepo is enough to proceed. Ask first only if the request is ambiguous or would widen the scope below.

2. Choose tasks

Configure only tasks whose behavior is clear, preferring lint, type-check, format-check, unit-test, and build tasks that the repository already runs.

Do not:

  • rename or move packages, or change toolchains, package managers, or workspace membership;
  • rewrite package-level task commands;
  • add Remote Caching;
  • configure deploy, end-to-end, integration, or environment-provisioning tasks unless asked.

Keep existing root entry points. Give the turbo version of a root orchestrator its own name following the repository's convention, such as <task>:turbo. Replace the original only if the turbo command is proven equivalent.

3. Add Turborepo

Install turbo as the toolchain's guide describes. If the repository pins its tool versions, pin turbo to an exact version the same way. Regenerate lockfiles with the tool that owns them. Never edit a lockfile by hand.

Add any prerequisites from the toolchain guide that are missing, such as a workspace name or toolchain declaration. Take versions from the lockfile or the toolchain's own command instead of guessing.

Create a root turbo.json for the selected tasks only:

  • Enable only the toolchain support those tasks need, including any required futureFlags. Tell the user if that support is experimental.
  • Add dependsOn, inputs, outputs, and environment variables only when the repository shows they are needed.
  • If you add custom inputs, include "$TURBO_DEFAULT$". Reference root files outside a package with "$TURBO_ROOT$/<path>".
  • Disable caching for side-effecting tasks and for tasks whose outputs cannot be captured.

Toolchain support can register built-in tasks you did not select. Report them as deferred, not validated.

Add the dependency directory and .turbo to .gitignore if they are missing. Do not change any other ignore rules.

4. Validate

Install dependencies in frozen or immutable mode. Then check a turbo run <task> --dry output for each task:

  • Only the intended packages and tasks are selected, and no package that defines a task is missing.
  • Task ordering matches what the repository does today.
  • For tasks with shared root inputs: run once, confirm a cache hit on the second run, change one root input, confirm a cache miss, then revert the change.

Then run each existing command and its turbo equivalent, and compare package coverage, ordering, exit status, and outputs. A passing turbo run alone does not prove equivalence. If behavior differs, narrow the configuration or stop. Never delete the old command to hide a difference.

5. Report

Summarize:

  • the configured tasks and package coverage;
  • preserved commands and any replaced root entry points;
  • inferred dependencies, inputs, outputs, and environment variables;
  • validation results comparing existing and turbo commands;
  • deferred or unsupported tasks.

Commit, push, or open a pull request only if the user asks. Open it as a draft.

When another skill calls this one

Work in the caller's branch, do not open a pull request, and return the diff, validation results, and deferred tasks to the caller. If the change would need to go beyond the selected tasks, stop instead of widening the scope.

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.