For large features or refactors, use plan mode to get Claude to think before it acts. Press Shift+Tab to toggle plan mode, or type /plan.
In plan mode, Claude operates with read-only tools — it can explore your codebase, read files, and research the problem space, but it won't write or modify anything until you give the go-ahead.
This is especially useful when building a new feature that touches multiple files, refactoring architecture, or when you want to review the approach before any code is written.
"I need to add two-factor authentication to the settings page.
Go into plan mode and figure out the best approach first."
Claude will explore the relevant files and lay out its proposed approach directly in the conversation. You can then discuss, refine, or redirect the plan before committing to anything.
Once you're happy with the direction, toggle plan mode off with Shift+Tab and Claude will begin implementing.
# A practical workflow
1. Describe a complex task
2. Shift+Tab → Claude explores and proposes an approach
3. Ask questions, push back, refine the plan in conversation
4. Shift+Tab again → Claude executes with confidence
Plan mode is where you catch misunderstandings — before they become messy diffs.
Log in to leave a comment.
The /security-review command scans your uncommitted changes for injection vectors, auth gaps, hardcoded secrets, and other common vulnerabilities.
The SessionStart hook fires when any session begins or resumes, making it ideal for loading environment variables and running one-time setup scripts.
Ask Claude to write property-based tests for your functions using fast-check — it identifies the mathematical invariants in your code and generates tests that cover inputs you'd never enumerate by hand.