Skip to main content
Context snippets are reusable text blocks that get injected into every Claude prompt for a project. Use them to enforce coding standards, architectural decisions, or project-specific rules.

​
How Snippets Work

When a Claude agent starts, all enabled snippets for the project are concatenated and included in the system prompt. Claude sees them as project rules it must follow.
1

Create a snippet

Go to project settings and add a context snippet with a name and content.
2

Enable or disable

Toggle snippets on or off. Disabled snippets are saved but not included in prompts.
3

Claude reads them

Every agent session includes all active snippets as context.

​
Example Snippets

Code Style Rules:
- Use TypeScript strict mode
- Prefer functional components with hooks
- Use named exports, not default exports
- Maximum file length: 300 lines

​
Enable/Disable Toggle

Each snippet has an independent toggle. This lets you:
  • Temporarily disable a rule without deleting it
  • Keep a library of snippets and activate only the relevant ones
  • A/B test different instructions across task runs
Snippets are different from CLAUDE.md. Snippets are managed in the UI and injected by Claude Board. CLAUDE.md is a file in your repo that Claude reads directly.

​
Managing Snippets

Create, edit, and delete snippets via project settings or the Snippets API.