claude-automation-recommender Skill (Anthropic) for Claude Code: What It Does and How to Install It

Makes Claude scan your project and recommend hooks, subagents, skills, plugins and MCP servers for your Claude Code setup.

Installs: 9.9K; on skills.sh
Stars: 45.7K; on GitHub
Files: 5; besides SKILL.md
Updated: Nov 2026; last repo commit

This skill looks at your codebase and tells you which Claude Code automations to add. It checks your language, frameworks, tests and CI. Then it recommends the top one or two hooks, subagents, skills, plugins and MCP servers, with a reason for each.

It is read-only: SKILL.md says it creates and modifies no files. It is Anthropic's own skill, from the claude-code-setup plugin. Ask for a single category to get 3 to 5 options.

Good for

Developers setting up Claude Code on a project for the first time
Teams standardizing hooks, agents and MCP servers
Anyone who wants to know which Claude Code features fit their stack

Install claude-automation-recommender

Claude Code

Copy this prompt.

Install the claude-automation-recommender skill in Claude Code. Run this command:

/plugin install claude-code-setup@claude-plugins-official

Then check that it loaded and tell me how to use it.

2
Paste it into Claude Code. It shows the change and asks before running it.

Pick one route. The plugin route installs claude-code-setup: one skill and no other skills, hooks, agents or MCP config, per the plugin folder in the repo tree. The skills-only routes install the same skill without the plugin wrapper.

Built from the package data we hold. Not yet tested by us.

What's in the skill

When it loads

Analyze a codebase and recommend Claude Code automations (hooks, subagents, skills, plugins, MCP servers). Use when user asks for automation recommendations, wants to optimize their Claude Code setup, mentions improving Claude Code workflows, asks how to first set up Claude Code for a project, or wants to know what Claude Code features they should use.

Sections in SKILL.md

Output Guidelines
Automation Types Overview
Workflow
Phase 1: Codebase Analysis
Phase 2: Generate Recommendations
Phase 3: Output Recommendations Report
Claude Code Automation Recommendations
Codebase Profile
🔌 MCP Servers
🎯 Skills
⚡ Hooks
🤖 Subagents
Decision Framework
When to Recommend MCP Servers
When to Recommend Skills
When to Recommend Hooks

Key facts

Author: Anthropic (the vendor's own repo: anthropics/claude-plugins-official; plugin README credits Isabella He)
License: Apache-2.0
Installs on skills.sh: 9,920 (Nov 2, 2026)
GitHub stars: 45,717 for the whole claude-plugins-official repo (Nov 2, 2026)
Last repo commit: Nov 1, 2026
Audits on skills.sh: Gen Agent Trust Hub: pass, Socket: warn, Snyk: pass
What it includes: 6 files including SKILL.md: 5 reference files in references/ (no scripts)
Part of: The claude-code-setup plugin in Anthropic's plugin marketplace (1 skill, per the plugin folder)
Allowed tools: Read, Glob, Grep, Bash

What it includes

Reads package.json, pyproject.toml, Cargo.toml, go.mod and the folder layout
Runs ls and cat style commands to find the language, dependencies, tests, CI and any existing .claude config.
Runs code · Phase 1: codebase analysis
Matches codebase signals to servers like context7, Playwright, Supabase, GitHub, Linear
Detail in references/mcp-servers.md.
Read · MCP server recommendations
Suggests existing plugin skills and custom skills to create
Examples: api-doc, create-migration, gen-test, pr-check, release-notes.
Read · Skill recommendations
Auto-format, auto-lint, type-check, run tests, block .env and lock file edits
Detail in references/hooks-patterns.md.
Read · Hook recommendations
Reviewers such as code-reviewer, security-reviewer, test-writer
Templates in references/subagent-templates.md.
Read · Subagent recommendations
Points to plugins like frontend-design and mcp-builder
Detail in references/plugins-reference.md.
Read · Plugin recommendations
Top 1-2 recommendations per category with why, where and how
Ends by telling you to ask for more of any category.
Read · Phase 3: report
MCP team sharing, headless claude -p examples, hook permissions
Shows sample claude -p commands for CI and pre-commit use.
Read · Configuration tips

From the official docs

Permissions

Low risk

What it can do

Read your data
Create and change things
Delete things
Read-only modeHow to turn it on

Watch out for

It uses Bash to inspect your project
The frontmatter allows Bash, and Phase 1 runs commands like ls and cat on config files. SKILL.md says the skill is read-only. Watch which commands run. github.com
Its advice installs things only if you act
It recommends MCP servers, hooks and plugins, with install commands in its report. Nothing is installed until you or Claude run them separately. Review each one first. github.com
It may search the web for ideas
SKILL.md tells Claude to use web search to find recommendations specific to your tools, frameworks and libraries, though web search is not in its allowed tools list. github.com

Security advisories

No published security advisories as of Nov 2, 2026.

Technical details

Turning on read-only	SKILL.md says the skill is read-only: it does not create or modify files.
Sign-in and keys	SKILL.md does not mention any account or key.
Scopes	Reads project files such as package.json, pyproject.toml and the folder layout Allowed tools: Read, Glob, Grep, Bash

Questions

What does the claude-automation-recommender skill do?

It makes Claude scan your codebase and recommend Claude Code setup: hooks, subagents, skills, plugins and MCP servers. It gives the top one or two per category with the reason for each.

Does it change my files?

SKILL.md says the skill is read-only. It analyzes the codebase and outputs recommendations, and does not create or modify files. You implement them yourself or ask Claude separately.

How do I install it?

Use the install box at the top of this page. It lists the claude-code-setup plugin from Anthropic's plugin marketplace, and skills-only routes. The skill runs shell commands such as ls and cat, so we do not offer a Claude app upload route.

Is this an Anthropic skill?

Yes. It is in anthropics/claude-plugins-official, in the plugin claude-code-setup, which lists Anthropic as author. That is Anthropic's own repo.

What does the plugin include?

The claude-code-setup plugin folder has this one skill with 5 reference files, a README, a license and a screenshot. The tree shows no hooks, agents or MCP config files in it.

What prompts trigger it?

The plugin README gives "recommend automations for this project", "help me set up Claude Code" and "what hooks should I use?".

Is it safe?

It bundles no scripts and its allowed tools are Read, Glob, Grep and Bash. It runs read-style shell commands on your project, and SKILL.md says it modifies nothing. Its recommendations include hooks and MCP servers, so review each before you add one.

Information

updated Nov 2, 2026

Repository: anthropics/claude-plugins-official
Folder: plugins/claude-code-setup/skills/claude-automation-recommender

anthropics/claude-plugins-official skills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 4

claude-plugins-official README · github.com
claude-code-setup README · github.com
claude-code-setup plugin.json · github.com
claude-automation-recommender SKILL.md · github.com