claude-guide

CLAUDE.md serves as the core configuration file for Claude Code automation. Organize it with modular folders: rules/ for governance, commands/ for quick shortcuts, skills/ for reusable packages, and agents/ for independent workflows. This structure determines load order and enables scalable, project-wide guidance while keeping core rules centralized and accessible.

Include in CLAUDE.md: core rules, environment variables, deployment steps, troubleshooting guidance, and service endpoints. Create separate skills for reusable, modular task packages and agents for independent workflows. Review your CLAUDE.md regularly to align with project needs and decide which sections to populate first based on team priorities.

Review CLAUDE.md by auditing its coverage against functional requirements: documented rules, quick commands, service endpoints, environment variables, deployment steps, and troubleshooting. Adapt sections to your project context, then decide which folders to populate first. This alignment streamlines Claude-based automation and ensures consistent governance across teams.

Yes. A well-structured CLAUDE.md with clear rules, commands, and deployment guidance accelerates onboarding by giving new team members instant access to project conventions, quick shortcuts, and troubleshooting steps. Organize modular skills and agents so contributors understand which components to reference for their tasks.

Rules define governance and constraints applied project-wide. Commands provide quick shortcuts for frequent tasks. Skills are reusable, modular packages solving specific problems independently. Agents are independent workflows orchestrating multiple actions. CLAUDE.md centralizes rules and core guidance; separate folders enable skill separation planning and scalable automation.

Create a new skill when you have a reusable, self-contained task package with clear inputs and outputs that serve multiple workflows. CLAUDE.md templates guide skill definition. Plan separation by identifying modular functions, ensuring each skill has independent components, and documenting how skills interact within your agents.