Skip to main content
skill CX Codex

conductor-setup — Claude Code skill: Conductor workspace setup

A skill that gets a repository ready for Conductor workspaces: it edits .conductor/settings.toml, migrates the old conductor.json, and sorts out scripts and env vars.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add biw/skills --skill conductor-setup -a claude-code
Open on GitHub

What conductor-setup does

Reach for this skill when a repository has to be prepared for Conductor workspaces. Called directly, it audits what is there against the conventions it carries and either makes the changes or tells you nothing needs doing.

The audit covers .conductor/settings.toml, .conductor/settings.local.toml, a legacy conductor.json, .worktreeinclude, .conductor/*.sh, root-level conductor-*.sh, package scripts and repo docs. From CONDUCTOR_IS_LOCAL it works out whether each script serves local workspaces, cloud ones, or both. Only the references the task needs get pulled in: settings and migration, scripts and environment, or copied files and troubleshooting.

Team settings win over machine-local ones, and hard-coded workspace paths, resources and local ports give way to Conductor variables. Secrets and per-machine credentials stay out of committed settings. MCP and privacy config is touched only when you ask or when repository policy requires it. TOML gets validated afterwards, and every script that changed is run through the narrowest relevant check.

What it helps with

  • move a repository onto .conductor/settings.toml without checking every field by hand
  • carry settings over from a legacy conductor.json
  • separate setup, run and archive scripts between local and cloud workspaces
  • drop hard-coded paths and local ports from the config
  • keep secrets and machine credentials out of commits

Example prompts for the agent

  • Set this repo up for Conductor workspaces and show me what has to change
  • Migrate the old conductor.json into .conductor/settings.toml
  • Check the scripts in .conductor/*.sh: which ones work in cloud workspaces and which are local only

How to install conductor-setup

1Skills CLI (Vercel)

npx skills add biw/skills --skill conductor-setup -a claude-code

2Manually: put the files in place

~/.claude/skills/conductor-setup/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/conductor-setup/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about conductor-setup

How do I install conductor-setup in Claude Code?

Run npx skills add biw/skills --skill conductor-setup -a claude-code. The skill file lands at ~/.claude/skills/conductor-setup/SKILL.md.

What does conductor-setup do?

It configures a repository for Conductor workspaces: edits .conductor/settings.toml, migrates the legacy conductor.json, and handles scripts, environment variables, copied files and caches. If the setup already matches, it reports that instead of changing anything.

How is it different from setting Conductor up by hand?

It checks the config against documented conventions and reads only the references the task crosses — settings and migration, scripts and environment, or copied files and diagnosis. TOML is validated after the edits, and each changed script gets the narrowest relevant check.

Source text

SKILL.md · MIT
---
name: conductor-setup
description: Configure .conductor/settings.toml, migrate legacy conductor.json, and set up local/cloud Conductor workspace scripts, env vars, files, and caches.
metadata:
  source: "biw/skills"
  homepage: "https://github.com/biw/skills"
---

# Conductor Setup

Use this skill when configuring a repository for Conductor workspaces. When invoked directly, audit the setup against the conventions below and either apply the changes or report that none are needed.

## Workflow

1. Inspect `.conductor/settings.toml`, `.conductor/settings.local.toml`, legacy `conductor.json`, `.worktreeinclude`, `.conductor/*.sh`, legacy root-level `conductor-*.sh`, package scripts, and repo docs. Check `CONDUCTOR_IS_LOCAL` and decide whether each script supports local workspaces, cloud workspaces, or both.
2. Read only the references needed for the task:
   - `references/settings-and-migration.md` for settings layers, schemas, supported repository fields, or `conductor.json` migration.
   - `references/scripts-and-environment.md` for setup/run/archive scripts, shells, variables, concurrency, Spotlight, or caches.
   - `references/files-layouts-and-troubleshooting.md` for Files to copy, `.worktreeinclude`, monorepos, linked repositories, MCP/privacy, or diagnosis.
     Read more than one only when the task crosses those concerns.
3. Apply the selected reference's documented contract. Prefer team settings over machine-local configuration; preserve an existing deliberate script layout; use Conductor variables instead of hard-coded workspace paths, resources, and local ports.
4. Keep secrets and machine-specific credentials out of committed settings. Change MCP/privacy configuration only when asked or required by repository policy.
5. Validate TOML and run the narrowest relevant check for every script changed. Report when the existing setup already satisfies the requested outcome.

## Resources

- `references/settings-and-migration.md`: settings scope, precedence, fields, schemas, and legacy migration.
- `references/scripts-and-environment.md`: script roles, local/cloud shells and variables, run behavior, caches, and cleanup.
- `references/files-layouts-and-troubleshooting.md`: copied files, repository layouts, privacy boundaries, and diagnosis.

© biw · MIT · biw/skills

Similar skills

All skills for Agent setup: memory, context, orchestration →

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor