Claude Automation Recommender

Claude Automation Recommender

Analyze codebase patterns to recommend tailored Claude Code automations across all extensibility options.

This skill is read-only. It analyzes the codebase and outputs recommendations. It does NOT create or modify any files. Users implement the recommendations themselves or ask Claude separately to help build them.

When to Use

Trigger this skill when the user:

• Asks for automation recommendations for their project
• Wants to optimize or improve their Claude Code setup
• Mentions improving Claude Code workflows
• Asks how to first set up Claude Code for a project
• Wants to know what Claude Code features they should use
• Asks about a specific automation type (hooks, subagents, skills, plugins, MCP servers)

Prerequisites

• Read access to the target codebase (project root)
• claude CLI installed and on PATH for any install/verification commands mentioned in recommendations
• On Windows host (primary), use PowerShell. Commands shown as ls -la are Bash equivalents; in PowerShell use Get-ChildItem -Force or dir /a if running natively. If running inside Claude Code's built-in Bash on Windows, the Bash commands work as-is.

Procedure

Phase 1: Codebase Analysis

Gather project context before generating any recommendations.

1. Detect project type and tooling.

# Detect project manifest files
ls -la package.json pyproject.toml Cargo.toml go.mod pom.xml 2>/dev/null

# Inspect package.json (first 50 lines)
cat package.json 2>/dev/null | head -50

PowerShell equivalent if running natively:

Get-ChildItem -Force package.json, pyproject.toml, Cargo.toml, go.mod, pom.xml -ErrorAction SilentlyContinue
Get-Content package.json -TotalCount 50 -ErrorAction SilentlyContinue

2. Check dependencies for MCP server recommendations.

cat package.json 2>/dev/null | grep -E '"(react|vue|angular|next|express|fastapi|django|prisma|supabase|convex|stripe)"'

3. Check for existing Claude Code configuration.

ls -la .claude/ CLAUDE.md 2>/dev/null

4. Analyze project structure.

ls -la src/ app/ lib/ tests/ components/ pages/ api/ 2>/dev/null

5. Capture key indicators using the table below.

Phase 2: Generate Recommendations

Based on analysis, generate recommendations across all categories. Only include 1-2 recommendations per category — the most valuable ones for this specific codebase. Skip categories that aren't relevant.

If the user asks for a specific type, focus only on that type and provide 3-5 recommendations instead.

A. MCP Server Recommendations

Load references/mcp-servers.md when generating MCP server recommendations for detailed patterns and install commands.

B. Skills Recommendations

Load references/skills-reference.md when generating skills recommendations for details on available skills and plugins.

Create skills in .claude/skills/<name>/SKILL.md. Some are also available via plugins:

Custom skills to create (with templates, scripts, examples):

Invocation control flags:

• disable-model-invocation: true — User-only (for side effects: deploy, commit, send)
• user-invocable: false — Claude-only (for background knowledge)
• Default (omit both) — Both can invoke

C. Hooks Recommendations

Load references/hooks-patterns.md when generating hooks recommendations for configurations.

D. Subagent Recommendations

Load references/subagent-templates.md when generating subagent recommendations for templates.

E. Plugin Recommendations

Load references/plugins-reference.md when generating plugin recommendations for available plugins.

Phase 3: Output Recommendations Report

Format recommendations using the template below. Go beyond the reference lists — use web search to find recommendations specific to the codebase's tools, frameworks, and libraries. End by noting the user can request more recommendations for any specific category.

## Claude Code Automation Recommendations

I've analyzed your codebase and identified the top automations for each category. Here are my top 1-2 recommendations per type:

### Codebase Profile
- **Type**: [detected language/runtime]
- **Framework**: [detected framework]
- **Key Libraries**: [relevant libraries detected]

---

### 🔌 MCP Servers

#### context7
**Why**: [specific reason based on detected libraries]
**Install**: `claude mcp add context7`

---

### 🎯 Skills

#### [skill name]
**Why**: [specific reason]
**Create**: `.claude/skills/[name]/SKILL.md`
**Invocation**: User-only / Both / Claude-only
**Also available in**: [plugin-name] plugin (if applicable)
```yaml
---
name: [skill-name]
description: [what it does]
disable-model-invocation: true  # for user-only
---

⚡ Hooks

[hook name]

Why: [specific reason based on detected config] Where: .claude/settings.json

🤖 Subagents

[agent name]

Why: [specific reason based on codebase patterns] Where: .claude/agents/[name].md

Want more? Ask for additional recommendations for any specific category (e.g., "show me more MCP server options" or "what other hooks would help?").

Want help implementing any of these? Just ask and I can help you set up any of the recommendations above.

## Decision Framework

### When to Recommend MCP Servers
- External service integration needed (databases, APIs)
- Documentation lookup for libraries/SDKs
- Browser automation or testing
- Team tool integration (GitHub, Linear, Slack)
- Cloud infrastructure management

### When to Recommend Skills
- Document generation (docx, xlsx, pptx, pdf — also in plugins)
- Frequently repeated prompts or workflows
- Project-specific tasks with arguments
- Applying templates or scripts to tasks (skills can bundle supporting files)
- Quick actions invoked with `/skill-name`
- Workflows that should run in isolation (`context: fork`)

### When to Recommend Hooks
- Repetitive post-edit actions (formatting, linting)
- Protection rules (block sensitive file edits)
- Validation checks (tests, type checks)

### When to Recommend Subagents
- Specialized expertise needed (security, performance)
- Parallel review workflows
- Background quality checks

### When to Recommend Plugins
- Need multiple related skills
- Want pre-packaged automation bundles
- Team-wide standardization

## Configuration Tips

### MCP Server Setup

**Team sharing**: Check `.mcp.json` into repo so entire team gets same MCP servers.

**Debugging**: Use `--mcp-debug` flag to identify configuration issues.

**Prerequisites to recommend:**
- GitHub CLI (`gh`) — enables native GitHub operations
- Puppeteer/Playwright CLI — for browser MCP servers

### Headless Mode (for CI/Automation)

Recommend headless Claude for automated pipelines:

```bash
# Pre-commit hook example
claude -p "fix lint errors in src/" --allowedTools Edit,Write

# CI pipeline with structured output
claude -p "<prompt>" --output-format stream-json | your_command

Permissions for Hooks

Configure allowed tools in .claude/settings.json:

{
  "permissions": {
    "allow": ["Edit", "Write", "Bash(npm test:*)", "Bash(git commit:*)"]
  }
}

Pitfalls

• Do NOT create or modify files. This skill is read-only. It analyzes and outputs recommendations only. Users implement them or ask Claude separately.
• Do not overwhelm with options. Default to 1-2 recommendations per category. Only expand to 3-5 when the user explicitly asks for a specific type.
• Do not skip the codebase analysis phase. Generic recommendations without codebase-specific reasoning are low value. Always tie "Why" to detected signals.
• Do not invent MCP server names or CLI flags. Only recommend servers that exist. Use references/mcp-servers.md for verified patterns.
• Do not forget invocation control flags. Skills with side effects (deploy, commit, send) should use disable-model-invocation: true. Background knowledge skills should use user-invocable: false.
• Windows path awareness: When recommending file paths in hooks or skill configs, use forward slashes in JSON (./.claude/settings.json) even on Windows — Claude Code config files use Unix-style paths.
• Do not recommend hooks that block the user's own edits without explanation. PreToolUse hooks that block .env or lock file edits should clearly state why in the recommendation output.

Verification

After generating recommendations, verify the analysis was thorough:

1. Confirm codebase profile accuracy.

# Re-check detected framework
cat package.json 2>/dev/null | grep -E '"(name|version|description)"'

2. Confirm existing Claude config was checked.

ls -la .claude/ CLAUDE.md .mcp.json 2>/dev/null

3. Confirm each recommendation has a codebase-specific "Why". If any recommendation's "Why" could apply to any project, it is too generic — revise it.

4. Confirm the output ends with the "Want more?" and "Want help implementing?" prompts so the user knows they can request deeper detail or implementation help.

Related Skills

• skill-development — For when the user wants to actually build a recommended skill
• writing-rules (hookify plugin) — For when the user wants to implement recommended hooks
• commit (commit-commands plugin) — For when the user wants to set up commit automation