| name | setup-ignore |
| description | Claude Code setup ignore: workflow guidelines, best practices, instructions, and integration examples |
Setup Ignore
When to activate
Activate when starting a new project session, when a user complains about high token usage/cost, or when Claude Code is reading too many irrelevant files (like package-lock.json, dist/, or vendor/). Invoked via /setup-ignore.
When NOT to use
Do not use if the project already has a well-configured .claudeignore file, or if the user explicitly needs Claude to analyze compiled output, minified files, or exact dependency lock versions.
Instructions
- Analyze the project structure to determine the primary language/framework (e.g., look for
package.json for Node, requirements.txt/pyproject.toml for Python, go.mod for Go).
- Fetch the corresponding template from
claudeignore-templates/ (e.g., node.claudeignore, python.claudeignore, go.claudeignore).
- If a
.claudeignore file already exists in the project root, intelligently merge the template rules into it, avoiding duplicates. Do not overwrite custom user rules.
- If a
.claudeignore file does not exist, create it in the project root using the template content.
- Explain to the user roughly how many tokens/files this will save from being needlessly indexed (e.g., "Ignored
node_modules and package-lock.json, which saves ~40k tokens per request").
Example
User: /setup-ignore
Claude: I see this is a Node.js project. I have created a .claudeignore file based on the Node template. I've ignored node_modules/, dist/, and package-lock.json. This will significantly reduce your context size and token costs on every turn.
