Every Claude Feature Ranked

One guide pulled from the video resource pack: the creator tier board, the S-tier countdown, the data-ranked Top 50, and the ranks 51-137 appendix.

Source Pack · What This Guide Contains

The downloaded deck is used only as a content source. This page uses the existing Samuelhuang guide design instead of the deck’s hand-drawn slide styling.

Methodology · How The Data Ranking Was Scored

The Markdown companion ranks for daily practical power. It is a separate data lens from the deck’s personal tier board.

Ranking process

• Discovery: 229 raw features across 10 categories, researched from official Claude Code docs and the official GitHub changelog.
• Verification: deduped to 142 distinct features, then confirmed 137 genuine Claude Code features.
• Judging: each feature scored 0-10 across five dimensions by a three-judge panel: daily power user, team lead, and strict skeptic.
• Dimensions: Impact, Breadth, Ease, Novelty, and Wow.

Categories: Agent SDK, Automation & Hooks, CI/CD, Code Quality, Context Management, Cost & Monitoring, IDE Integrations, Integrations, Memory, Model Capabilities, Non-Interactive Mode, Parallel Work, Permissions & Security, Plugins, Project Setup, Research, Session Management, Skills, Sub-Agents, Surfaces & Access, Workflow Control.

Composite weights

Video Tier Board · The Deck’s Claude Ecosystem Ranking

The board below preserves the deck’s broader Claude ecosystem list. Hot cards were called out in the original deck data.

A star (★) marks the cards the original deck flagged as highlights.

Tier S · The top 12

The features the deck says changed day-to-day work.

Tier A · Honorable mentions

Strong Claude ecosystem features that sit just below S.

Tier B · Genuinely useful

Good features that are not the personal top tier.

Tier C · Useful, but lower for this workflow

Cowork, web, and interaction features that matter less to this ranking lens.

Tier D · Table stakes and plumbing

Important basics that are real and useful, but rarely favorite features.

Data Ranking · The Top 50 Claude Code Features

Each entry keeps the feature description, practical value, score breakdown, category, composite score, and official source URL from the Markdown companion.

#1 · /plan (Plan Mode)

Workflow Control · Composite 8.03 · Impact 8.5 · Breadth 8.2 · Ease 8.5 · Novelty 6.3 · Wow 7.0

What it does: Switches Claude to read-only plan mode: it maps the codebase and writes a plan but makes zero edits until you approve.

Why it matters: Catch a bad direction before Claude writes a hundred lines of wrong code.

Source: https://code.claude.com/docs/en/commands

#2 · /rewind (Checkpointing)

Workflow Control · Composite 7.99 · Impact 8.2 · Breadth 7.7 · Ease 8.8 · Novelty 7.3 · Wow 7.7

What it does: Claude auto-snapshots all file edits before each prompt; run /rewind or press Esc twice on an empty prompt to restore code, conversation, or both from any prior checkpoint.

Why it matters: Fearless experimentation: every prompt is a save point you can roll back to.

Source: https://code.claude.com/docs/en/checkpointing

#3 · @-File Mentions

Context Management · Composite 7.58 · Impact 7.7 · Breadth 8.8 · Ease 9.3 · Novelty 4.2 · Wow 4.3

What it does: Type @ followed by a file or directory path in your prompt to trigger autocomplete and inject its contents directly into the prompt.

Why it matters: Gives Claude exactly the context it needs without waiting for it to discover files on its own.

Source: https://code.claude.com/docs/en/interactive-mode

#4 · CLAUDE.md Memory Hierarchy

Memory · Composite 7.47 · Impact 8.3 · Breadth 8.3 · Ease 7.2 · Novelty 5.5 · Wow 4.3

What it does: CLAUDE.md loads from four scopes every session: managed policy, user (~/.claude), project (./CLAUDE.md or .claude/CLAUDE.md), and local (CLAUDE.local.md), all concatenated in order.

Why it matters: You can layer org-wide rules, personal preferences, and project conventions without conflicts.

Source: https://code.claude.com/docs/en/memory

#5 · Auto-Compaction

Context Management · Composite 7.45 · Impact 7.0 · Breadth 8.5 · Ease 10.0 · Novelty 4.8 · Wow 4.7

What it does: When a session nears its context limit, Claude Code automatically summarizes history and re-injects CLAUDE.md so the session keeps running without you doing anything.

Why it matters: Long tasks don’t die at the context wall; they continue automatically.

Source: https://code.claude.com/docs/en/costs

#6 · Skills (SKILL.md Custom Commands)

Skills · Composite 7.35 · Impact 8.5 · Breadth 6.5 · Ease 6.5 · Novelty 7.2 · Wow 7.3

What it does: Markdown instruction files in .claude/skills/<name>/ that Claude loads on demand when invoked or relevant, following the open Agent Skills standard.

Why it matters: Package repeatable workflows as reusable commands without bloating the always-on system prompt.

Source: https://code.claude.com/docs/en/skills

#7 · ESC to Interrupt

Workflow Control · Composite 7.34 · Impact 7.2 · Breadth 8.8 · Ease 10.0 · Novelty 3.2 · Wow 3.7

What it does: Press Esc during any response or tool call to stop Claude mid-turn while preserving work it has already completed.

Why it matters: Redirect immediately when Claude heads the wrong direction without losing what it already did right.

Source: https://code.claude.com/docs/en/interactive-mode

#8 · Shift+Tab Permission Mode Cycling

Workflow Control · Composite 7.28 · Impact 7.3 · Breadth 7.7 · Ease 9.0 · Novelty 5.7 · Wow 5.0

What it does: Press Shift+Tab to cycle permission modes (default, acceptEdits, plan, and any extras you have enabled such as auto or bypassPermissions) without restarting the session.

Why it matters: Switch from careful mode to auto-accept mid-session when you trust the direction Claude is heading.

Source: https://code.claude.com/docs/en/interactive-mode

#9 · Built-in Subagents (Explore, Plan, General-purpose)

Sub-Agents · Composite 7.23 · Impact 7.2 · Breadth 7.2 · Ease 9.3 · Novelty 6.2 · Wow 5.5

What it does: Three pre-configured subagents ship by default: Explore (read-only Haiku for search), Plan (read-only, used in plan mode), and General-purpose (all tools, for complex multi-step tasks).

Why it matters: Out of the box, Claude already routes search and planning to cost-optimized workers.

Source: https://code.claude.com/docs/en/sub-agents

#10 · /code-review

Code Quality · Composite 7.20 · Impact 7.7 · Breadth 6.8 · Ease 8.3 · Novelty 5.7 · Wow 6.5

What it does: Bundled skill that reviews the current diff for correctness bugs and cleanups at a selectable effort level, with --fix to auto-apply findings and --comment to post inline GitHub PR comments.

Why it matters: Automated code review on every diff with one command, with findings applied or posted without leaving the terminal.

Source: https://code.claude.com/docs/en/commands

#11 · /clear

Context Management · Composite 7.10 · Impact 6.7 · Breadth 9.0 · Ease 10.0 · Novelty 2.8 · Wow 2.8

What it does: Starts a new conversation with empty context while keeping the previous one available to return to via /resume.

Why it matters: Clean slate for a new task without losing the thread of earlier work.

Source: https://code.claude.com/docs/en/commands

#12 · Permission Modes (default / acceptEdits / plan / auto / dontAsk / bypassPermissions)

Permissions & Security · Composite 7.10 · Impact 7.5 · Breadth 7.8 · Ease 7.5 · Novelty 5.2 · Wow 4.8

What it does: Six modes (default, acceptEdits, plan, auto, dontAsk, bypassPermissions) control how aggressively Claude acts without prompting, from asking before every tool call to skipping all checks.

Why it matters: You dial autonomy precisely: high trust for personal machines, tight controls for shared or sensitive repos.

Source: https://code.claude.com/docs/en/settings

#13 · opusplan Model Alias

Workflow Control · Composite 6.99 · Impact 7.3 · Breadth 6.2 · Ease 7.8 · Novelty 7.5 · Wow 6.5

What it does: Set your model to opusplan and Claude uses Opus during plan mode for reasoning, then automatically switches to Sonnet for execution.

Why it matters: You get heavy reasoning where it counts and cheaper generation where it doesn’t.

Source: https://code.claude.com/docs/en/model-config

#14 · Auto Memory (MEMORY.md)

Memory · Composite 6.94 · Impact 6.8 · Breadth 6.5 · Ease 8.7 · Novelty 7.2 · Wow 5.8

What it does: Claude automatically writes build commands, debugging insights, and preferences it discovers into ~/.claude/projects/<repo>/memory/MEMORY.md; the first 200 lines load into every session automatically.

Why it matters: Claude gets smarter about your project over time without you manually updating CLAUDE.md.

Source: https://code.claude.com/docs/en/memory

#15 · ultrathink Keyword

Workflow Control · Composite 6.92 · Impact 6.5 · Breadth 6.8 · Ease 9.3 · Novelty 6.2 · Wow 5.8

What it does: Include ultrathink anywhere in a prompt to trigger deeper reasoning on that single turn without changing your session effort setting. No other keyword phrases (think hard, think more) are recognized.

Why it matters: One-off deep reasoning on demand without touching your effort configuration.

Source: https://code.claude.com/docs/en/model-config

#16 · /init

Project Setup · Composite 6.91 · Impact 6.8 · Breadth 7.5 · Ease 9.0 · Novelty 4.3 · Wow 4.8

What it does: Generates a starter CLAUDE.md by analyzing your codebase; set CLAUDE_CODE_NEW_INIT=1 to enable an interactive multi-phase flow that also walks through skills, hooks, and memory files.

Why it matters: Cuts setup time for new repos to seconds instead of writing CLAUDE.md from scratch.

Source: https://code.claude.com/docs/en/commands

#17 · /compact

Context Management · Composite 6.91 · Impact 6.8 · Breadth 7.7 · Ease 9.0 · Novelty 4.7 · Wow 4.0

What it does: Summarizes the current conversation to free context, with an optional focus instruction (e.g. /compact focus on code samples) to shape what gets preserved.

Why it matters: Extends long sessions without losing the plot on what matters.

Source: https://code.claude.com/docs/en/commands

#18 · Task List (TodoWrite / Ctrl+T)

Workflow Control · Composite 6.91 · Impact 6.3 · Breadth 7.8 · Ease 9.2 · Novelty 4.5 · Wow 5.2

What it does: For multi-step work Claude maintains a task list in the terminal status area showing pending, in-progress, and completed items; toggle visibility with Ctrl+T.

Why it matters: Track where Claude is in a long task at a glance without reading the transcript.

Source: https://code.claude.com/docs/en/interactive-mode

#19 · MCP Tool Search (Deferred Loading)

Integrations · Composite 6.91 · Impact 7.0 · Breadth 6.0 · Ease 9.2 · Novelty 7.3 · Wow 5.5

What it does: Enabled by default: only tool names load at session start and Claude fetches full schemas on demand via a ToolSearch call, so large MCP server collections barely affect context.

Why it matters: Lets you connect many MCP servers without burning context window space on tools Claude never uses.

Source: https://code.claude.com/docs/en/mcp

#20 · Custom Subagents (.claude/agents/)

Sub-Agents · Composite 6.84 · Impact 7.7 · Breadth 6.3 · Ease 6.5 · Novelty 6.3 · Wow 6.5

What it does: Define reusable specialist agents as markdown files with YAML frontmatter in .claude/agents/ or ~/.claude/agents/ that Claude auto-delegates to based on their description field.

Why it matters: Build a stable of purpose-built workers (researcher, code-reviewer, etc.) that activate automatically when relevant.

Source: https://code.claude.com/docs/en/sub-agents

#21 · /context

Context Management · Composite 6.82 · Impact 6.3 · Breadth 6.5 · Ease 9.2 · Novelty 6.5 · Wow 6.3

What it does: Visualizes context window usage as a colored grid and surfaces optimization suggestions for memory bloat and capacity warnings.

Why it matters: You can see exactly what is eating your context before it becomes a problem.

Source: https://code.claude.com/docs/en/commands

#22 · VS Code Extension

IDE Integrations · Composite 6.81 · Impact 7.5 · Breadth 5.8 · Ease 7.0 · Novelty 6.3 · Wow 7.5

What it does: Native VS Code panel with side-by-side inline diff review before any file is written, @-mention with line-range context, an editable Markdown plan in Plan mode, and per-message checkpoint rewind.

Why it matters: Review and control Claude’s file changes visually inside VS Code without switching to a terminal.

Source: https://code.claude.com/docs/en/vs-code

#23 · /security-review

Code Quality · Composite 6.73 · Impact 7.0 · Breadth 5.8 · Ease 8.3 · Novelty 6.3 · Wow 6.5

What it does: Built-in command that analyzes the current git diff for security vulnerabilities including injection, auth issues, and data exposure.

Why it matters: Catches security regressions before they leave the branch, without leaving the terminal.

Source: https://code.claude.com/docs/en/commands

#24 · /verify

Code Quality · Composite 6.68 · Impact 7.0 · Breadth 6.0 · Ease 7.5 · Novelty 6.5 · Wow 6.5

What it does: Bundled skill that builds and runs your app, then watches the result to confirm a code change works instead of relying on tests or type checks.

Why it matters: Validates changes against live app behavior, which catches runtime issues that static analysis misses.

Source: https://code.claude.com/docs/en/skills

#25 · 1 Million Token Context Window

Model Capabilities · Composite 6.68 · Impact 7.2 · Breadth 5.3 · Ease 8.2 · Novelty 6.3 · Wow 7.2

What it does: Opus and Sonnet support a 1 million token context window via the opus[1m] or sonnet[1m] model aliases, or by appending [1m] to a full model name.

Why it matters: Handle massive codebases or long sessions without hitting context limits on complex multi-file tasks.

Source: https://code.claude.com/docs/en/model-config

#26 · /deep-research

Research · Composite 6.65 · Impact 6.7 · Breadth 5.8 · Ease 8.3 · Novelty 6.0 · Wow 7.2

What it does: Bundled workflow that fans out web searches across angles, fetches and cross-checks sources, and synthesizes a cited report.

Why it matters: Delivers a sourced research brief in one command instead of manually searching and reading multiple pages.

Source: https://code.claude.com/docs/en/commands

#27 · /resume

Workflow Control · Composite 6.61 · Impact 6.7 · Breadth 7.2 · Ease 8.8 · Novelty 4.0 · Wow 4.0

What it does: Reopens a previous conversation by ID, name, or interactive picker; background sessions appear in the picker marked with bg.

Why it matters: Pick up any past session exactly where you left off, including sessions that ran in the background.

Source: https://code.claude.com/docs/en/commands

#28 · /mcp

Integrations · Composite 6.59 · Impact 7.2 · Breadth 6.7 · Ease 7.2 · Novelty 5.0 · Wow 5.0

What it does: Manages MCP server connections and OAuth authentication for external tool integrations.

Why it matters: Single command to connect, inspect, and authenticate any MCP server without editing config files.

Source: https://code.claude.com/docs/en/commands

#29 · GitHub Actions (@claude Trigger)

CI/CD · Composite 6.53 · Impact 7.5 · Breadth 5.7 · Ease 5.7 · Novelty 6.3 · Wow 7.2

What it does: Install the Claude GitHub App and add a workflow YAML to get @claude mentions in PRs and issues, plus event-driven and scheduled automation with full CLI flag passthrough via claude_args.

Why it matters: Turns GitHub comments into automated code changes and PRs without leaving your normal review workflow.

Source: https://code.claude.com/docs/en/github-actions

#30 · Auto Mode with ML Classifier

Permissions & Security · Composite 6.50 · Impact 6.8 · Breadth 5.3 · Ease 7.2 · Novelty 7.8 · Wow 6.5

What it does: In auto mode, a built-in ML classifier evaluates each tool call against environment context and custom allow/soft-deny/hard-deny rule sets to decide whether to approve, prompt, or block.

Why it matters: Reduces interruptions on routine operations while still catching genuinely risky actions.

Source: https://code.claude.com/docs/en/settings

#31 · Hooks System

Automation & Hooks · Composite 6.50 · Impact 7.8 · Breadth 5.5 · Ease 4.8 · Novelty 7.0 · Wow 6.8

What it does: User-defined shell commands, HTTP endpoints, LLM prompt evaluations, MCP tools, or spawned subagents that run automatically at specific lifecycle points (PreToolUse, PostToolUse, Notification, etc.).

Why it matters: Adds deterministic enforcement and automation that doesn’t depend on Claude choosing to act.

Source: https://code.claude.com/docs/en/hooks-guide

#32 · --print / -p (Headless Mode)

Non-Interactive Mode · Composite 6.46 · Impact 7.3 · Breadth 5.7 · Ease 7.5 · Novelty 5.2 · Wow 5.5

What it does: Runs Claude non-interactively: sends a prompt, prints the response, and exits; accepts piped stdin content.

Why it matters: Compose Claude into shell pipelines, CI scripts, and automation without any interactive UI.

Source: https://code.claude.com/docs/en/cli-reference

#33 · /model

Workflow Control · Composite 6.42 · Impact 6.0 · Breadth 7.3 · Ease 9.3 · Novelty 3.5 · Wow 3.7

What it does: Switches the model mid-session; as of v2.1.153 it also saves the choice as your default for new sessions. The picker includes a left/right arrow effort-level slider for supported models.

Why it matters: Swap to Opus for a hard problem and back to Sonnet for routine work without restarting.

Source: https://code.claude.com/docs/en/commands

#34 · /usage

Cost & Monitoring · Composite 6.42 · Impact 5.7 · Breadth 7.2 · Ease 9.3 · Novelty 4.3 · Wow 4.5

What it does: Shows session cost, plan usage limits, and activity stats broken down by skill, subagent, plugin, and MCP server. /cost and /stats are aliases.

Why it matters: Pinpoints which skills or agents are burning the most tokens so you can optimize session cost.

Source: https://code.claude.com/docs/en/commands

#35 · Claude Code on the Web

Surfaces & Access · Composite 6.42 · Impact 6.8 · Breadth 5.2 · Ease 7.0 · Novelty 6.8 · Wow 7.5

What it does: A browser-based surface at claude.ai/code that runs sessions on Anthropic-managed cloud infrastructure, clones GitHub repos, and keeps sessions alive after the browser tab closes.

Why it matters: Kick off long-running tasks without a local setup and check back from any device when they finish.

Source: https://code.claude.com/docs/en/claude-code-on-the-web

#36 · Image Paste (Ctrl+V)

Workflow Control · Composite 6.38 · Impact 6.3 · Breadth 6.5 · Ease 8.3 · Novelty 4.3 · Wow 5.3

What it does: Press Ctrl+V (or Cmd+V in iTerm2, Alt+V on Windows/WSL) to paste a clipboard image into the prompt as an [Image #N] chip you can reference in your message.

Why it matters: Drop in a screenshot of a bug or a design mockup without leaving the terminal.

Source: https://code.claude.com/docs/en/interactive-mode

#37 · Remote Control

Session Management · Composite 6.38 · Impact 6.7 · Breadth 4.8 · Ease 6.7 · Novelty 7.7 · Wow 8.3

What it does: Connects claude.ai/code or the Claude mobile app to a local Claude Code session so you can drive it from any browser or phone, including sending messages and approving permission prompts remotely.

Why it matters: Continue in-progress local work from a phone or another computer without moving files to the cloud.

Source: https://code.claude.com/docs/en/remote-control

#38 · Allow / Deny / Ask Permission Rules

Permissions & Security · Composite 6.32 · Impact 7.2 · Breadth 6.7 · Ease 6.5 · Novelty 4.5 · Wow 3.8

What it does: Granular settings.json rules whitelist or block specific tool calls (e.g., Bash(npm run *), Read(.env), WebFetch) with deny evaluated first before ask, then allow.

Why it matters: Prevents accidental destructive commands or credential reads without blocking everything else.

Source: https://code.claude.com/docs/en/settings

#39 · @-Import Syntax in CLAUDE.md

Memory · Composite 6.29 · Impact 7.0 · Breadth 6.0 · Ease 6.8 · Novelty 5.7 · Wow 4.5

What it does: Use @path/to/file inside any CLAUDE.md to pull in external docs at session start; imported files load fully into context and can recursively import others up to four hops deep.

Why it matters: Keeps CLAUDE.md concise by referencing shared docs like READMEs and package.json rather than duplicating them.

Source: https://code.claude.com/docs/en/memory

#40 · /permissions

Permissions & Security · Composite 6.23 · Impact 6.3 · Breadth 6.7 · Ease 8.2 · Novelty 4.2 · Wow 3.7

What it does: Interactive dialog to manage allow, ask, and deny rules by scope, including review of recent auto-mode denials.

Why it matters: Gives you a single UI to tune Claude’s permissions without editing settings JSON by hand.

Source: https://code.claude.com/docs/en/commands

#41 · VS Code @browser Chrome Integration

IDE Integrations · Composite 6.23 · Impact 6.8 · Breadth 4.7 · Ease 5.8 · Novelty 7.5 · Wow 8.2

What it does: Type @browser in the VS Code prompt to give Claude control of your Chrome browser via the Claude in Chrome extension: open tabs, read console errors, and automate web workflows using your existing login state.

Why it matters: Debug live web apps and automate browser tasks without leaving VS Code or managing separate browser automation setup.

Source: https://code.claude.com/docs/en/vs-code

#42 · Plugin Marketplaces

Plugins · Composite 6.21 · Impact 6.5 · Breadth 5.2 · Ease 6.7 · Novelty 6.7 · Wow 7.2

What it does: Host a marketplace.json catalog in any git repo and users install plugins via /plugin install; Anthropic maintains an official curated marketplace and a community marketplace for third-party submissions.

Why it matters: Teams can self-host internal plugin catalogs while the community marketplace gives access to vetted third-party plugins.

Source: https://code.claude.com/docs/en/plugin-marketplaces

#43 · /code-review ultra

Code Quality · Composite 6.18 · Impact 7.0 · Breadth 4.2 · Ease 6.7 · Novelty 7.3 · Wow 7.5

What it does: Running /code-review ultra (or the alias /ultrareview) runs a deep multi-agent code review in a cloud sandbox.

Why it matters: Thorough multi-agent review for high-stakes PRs without spinning up your own infrastructure.

Source: https://code.claude.com/docs/en/commands

#44 · Extended Thinking Toggle (Alt+T)

Workflow Control · Composite 6.17 · Impact 6.2 · Breadth 5.8 · Ease 8.7 · Novelty 4.8 · Wow 4.8

What it does: Press Option+T (macOS) or Alt+T (Windows/Linux) to toggle extended thinking for the session; on supported models Claude also reasons between tool calls (interleaved thinking).

Why it matters: Turn on deep reasoning globally for a session when working through a complex architectural problem.

Source: https://code.claude.com/docs/en/model-config

#45 · /effort

Workflow Control · Composite 6.16 · Impact 6.2 · Breadth 5.5 · Ease 8.3 · Novelty 6.0 · Wow 5.0

What it does: Sets reasoning effort (low, medium, high, xhigh, max, ultracode) via an interactive slider or direct argument, controlling how much the model thinks per turn.

Why it matters: Trade cost for capability on a per-task basis without changing your model.

Source: https://code.claude.com/docs/en/commands

#46 · /batch

Parallel Work · Composite 6.14 · Impact 7.2 · Breadth 3.8 · Ease 5.7 · Novelty 8.2 · Wow 8.2

What it does: Bundled skill that decomposes a large change into 5 to 30 independent units, presents a plan for approval, then spawns one background subagent per unit in an isolated git worktree.

Why it matters: Parallelize a codebase-wide migration or refactor that would take hours done sequentially.

Source: https://code.claude.com/docs/en/commands

#47 · .mcp.json Project-Scoped Servers

Integrations · Composite 6.11 · Impact 6.8 · Breadth 5.8 · Ease 6.7 · Novelty 5.3 · Wow 4.3

What it does: A .mcp.json file at the project root defines MCP servers shared via version control, with ${VAR} and ${VAR:-default} environment-variable expansion in command, args, env, url, and headers.

Why it matters: Lets teams share MCP server configs in git while keeping secrets out of the repo via env var references.

Source: https://code.claude.com/docs/en/mcp

#48 · Agent SDK (Python & TypeScript)

Agent SDK · Composite 6.11 · Impact 7.5 · Breadth 4.3 · Ease 5.3 · Novelty 6.5 · Wow 7.3

What it does: The Claude Agent SDK exposes the same agent loop as Claude Code as a programmable async streaming API in both Python and TypeScript, with built-in tools for file read/edit, Bash, and web search.

Why it matters: Embed a full Claude Code agent directly into your application without shelling out to the CLI.

Source: https://code.claude.com/docs/en/agent-sdk/overview

#49 · /fewer-permission-prompts

Permissions & Security · Composite 6.10 · Impact 5.8 · Breadth 5.5 · Ease 8.2 · Novelty 6.5 · Wow 5.3

What it does: Bundled skill that scans transcripts for common read-only Bash and MCP tool calls and writes a prioritized allowlist to project settings to reduce permission prompts.

Why it matters: Cuts repetitive permission dialogs for safe, routine operations across a project.

Source: https://code.claude.com/docs/en/commands

#50 · /simplify

Code Quality · Composite 6.07 · Impact 6.2 · Breadth 5.0 · Ease 8.2 · Novelty 6.0 · Wow 5.8

What it does: Bundled skill that runs four parallel review agents covering reuse, simplification, efficiency, and abstraction level, then applies fixes. It does not hunt for correctness bugs.

Why it matters: Cleans up code quality in one command without mixing cleanup concerns with bug-finding.

Source: https://code.claude.com/docs/en/commands

Appendix · Ranks 51-137

The remaining confirmed Claude Code features from the companion ranking.

Every Claude Feature Ranked. Built from the local resource pack for Samuelhuang.