Claude Code

1. Claude Code overview

   What Claude Code is: agentic coding in terminal and IDE, the agent loop, what it can read, edit and run.

2. Common workflows

   Everyday usage patterns: exploring a codebase, plan mode, fixing bugs, writing tests, git workflows, working with images, resuming sessions.

3. Memory and CLAUDE.md

   CLAUDE.md files and their hierarchy (user, project, local), imports, auto memory, persistent project instructions.

4. Settings files and precedence

   settings.json at user, project and local scope, enterprise managed settings, precedence order, environment variables and configuration keys.

5. Permissions

   Allow, ask and deny rules for tools, rule syntax for Bash/Read/Edit/WebFetch, directories, tool permission evaluation.

6. Permission modes

   Default, acceptEdits, plan, auto and bypassPermissions modes, how each changes prompting and autonomy.

7. Sandboxing

   Sandboxed Bash tool with filesystem and network isolation, allowed domains, reducing permission prompts safely.

8. Hooks

   Lifecycle hooks (PreToolUse, PostToolUse, UserPromptSubmit, Stop, SessionStart), matchers, command hooks, JSON input/output, exit codes that block actions.

9. Subagents

   Custom subagents defined as markdown with frontmatter, separate context windows, tool restrictions, delegation and parallel agents.

10. Skills

    SKILL.md packaged instructions, progressive disclosure, frontmatter, scripts and resources, automatic and manual invocation.

11. Commands and slash commands

    Built-in slash commands, custom commands as markdown files, arguments, project vs user commands.

12. MCP servers in Claude Code

    Adding MCP servers (stdio, HTTP, SSE), scopes, OAuth authentication, using MCP tools, resources and prompts, .mcp.json.

13. Plugins

    Packaging commands, agents, skills, hooks and MCP servers as plugins, marketplaces and distribution.

14. Checkpointing

    Automatic checkpoints of file edits, rewinding code and conversation state, limitations with bash changes.

15. Model configuration

    Selecting models and aliases, opusplan, effort and thinking settings, environment variables for model overrides.

16. Headless mode and programmatic use

    Running claude -p non-interactively, output formats (text, json, stream-json), session resumption, scripting and CI usage.

17. Agent SDK

    Building custom agents with the Claude Agent SDK in Python and TypeScript, query loop, custom tools, hooks and permissions in code.

18. GitHub Actions integration

    Running Claude Code in CI, @claude mentions on PRs and issues, action configuration and secrets.

19. Monitoring and usage

    OpenTelemetry metrics and events, token and cost tracking, telemetry configuration for teams.

Each section links to the official page. Titles and summaries by Extra Credit; the docs belong to their authors.

Source: endoflife.date