Agents and Plan Mode
How Claude Code separates planning, coordination, and delegated execution
Understand the difference between the main thread, plan mode, coordinator behavior, and constrained sub-agents.
Plan mode is a workflow boundary, not just a UI toggle
Plan mode changes the permission context and is treated as a separate collaboration state.
The goal is to align on intent and implementation before the assistant starts mutating the codebase.
- The plan can be shown, stored, and reopened in an editor.
- Plan mode has dedicated enter and exit behaviors.
- This state is part of how Claude Code manages non-trivial tasks.
The main thread and delegated agents have different responsibilities
The main thread keeps control over the user-facing workflow, permission-sensitive actions, and coordination state.
Sub-agents are narrower workers. They are useful for bounded work, not for inheriting the full conversation brain.
- Some tools are disallowed entirely for async agents.
- Coordinator mode focuses on orchestration rather than direct file work.
- Task and messaging tools are treated differently for in-process teammates.
Good delegation is constrained delegation
The tool restrictions make the design philosophy visible: delegate bounded, concrete work and keep global state changes close to the main thread.
That is why plan state, recursive control, and some session-level behaviors stay centralized.
Learn next
These guides are adjacent to the current topic and make the learning path easier to follow.
Context and Memory
How Claude Code builds working context and persistent memory
Understand the difference between session context, CLAUDE.md instructions, MEMORY.md indexing, and relevance-based memory recall.
How Claude Code Works
The high-level architecture behind a Claude Code turn
A code-backed explanation of how Claude Code assembles context, runs the agent loop, invokes tools, and keeps long sessions moving.
MCP and Plugins
How Claude Code connects to external servers, tools, and resources
A map of MCP connection management, plugin-facing flows, and how external capabilities become part of the Claude Code experience.
Related resources
These directory entries connect the concept guide to concrete tools, docs, and integrations.
Claude Code Best Practices
Community-driven best practices for Claude Code
Collection of best practices learned from real-world Claude Code usage. Covers CLAUDE.md optimization, effective prompting, workflow organization, and common pitfalls to avoid. Continuously updated by the community.
Code Review Prompts
Structured prompts for thorough code reviews
Collection of prompts for conducting comprehensive code reviews. Covers security analysis, performance review, best practices, error handling, and maintainability assessment. Designed for systematic code evaluation.
Refactoring Guide Prompts
Prompts for safe and systematic code refactoring
Prompts to guide code refactoring while maintaining functionality. Covers identifying refactoring opportunities, planning changes, ensuring test coverage, and validating behavior preservation.
Sequential Thinking MCP Server
Structured step-by-step reasoning and problem decomposition
MCP server that enables structured sequential thinking and reasoning. Break down complex problems into manageable steps, maintain thought chains, and improve Claude's problem-solving capabilities with explicit reasoning traces.