SLOPSHOPPER

agent4everything

Maps .agents/skills (global and per project) into .claude/skills so every skill is active, without duplicating existing ones

newtoastprocess
★ 230v0.1.0no licenseupdated 2026-10-07Melvynx/aiblueprint/agents-config/claude-config/mods/agent4everything
A shopper browsing a rack in a slop shop
README

AIBlueprint CLI

Supercharge Claude Code and Codex with security hooks, reusable skills, a lightweight statusline, and workflow automation.

📚 Full Documentation | 🎯 Premium Features

🚀 Quick Start

# Run setup (no installation required)
npx aiblueprint-cli@latest agents setup

✨ Features

  • 🛡️ Security - Command validation hooks blocking dangerous operations
  • 📊 Statusline - Git status and session context display
  • 🤖 Skills - 30+ reusable workflow skills
  • 🎭 Agents - 3 specialized AI agents for codebase exploration
  • ⚡ Scripts - Built-in statusline utilities
  • 🧩 Mods - Claude Code function-hook mods (usage band, skills sync)

📦 Installation

Option 1: Plugin (Recommended)

/plugin marketplace add melvynx/aiblueprint
/plugin install aibp-base@AIBlueprint

Option 2: CLI Tool

# Run without installation
npx aiblueprint-cli@latest agents setup

🎯 Usage

Setup

# Interactive setup
npx aiblueprint-cli@latest agents setup

# Install all features (no prompts)
npx aiblueprint-cli@latest agents setup --skip

# Custom location
npx aiblueprint-cli@latest agents setup --folder ~/.my-claude

Statusline

Install a lightweight Claude Code statusline:

npx aiblueprint-cli@latest agents setup

Other Commands

# Add specific hooks
npx aiblueprint-cli@latest agents add hook post-edit-typescript

# Create symlinks between tools
npx aiblueprint-cli@latest agents symlink

# Centralize global skills and agents in ~/.agents
npx aiblueprint-cli@latest agents unify

# Choose what to centralize interactively
npx aiblueprint-cli@latest agents unify -i

# Unify project-local .claude/.cursor config into .agents
npx aiblueprint-cli@latest agents unify projects

# Unify only selected project categories
npx aiblueprint-cli@latest agents unify projects --agents-md --skills --rules

# Recover sessions from saved configs and backups
npx aiblueprint-cli@latest agents config unify sessions

📚 What You Get

Security Hooks

  • Command validation - Blocks rm -rf, privilege escalation, remote execution
  • TypeScript processing - Auto-format and lint after file edits
  • Security logging - Tracks all blocked commands to ~/.claude/security.log

Custom Statusline

  • Git branch, changes, and repository info
  • Session context and duration
  • Colored visual indicators

Workflow Skills

The free configuration ships the skills listed in the catalog below. The CLI installs every directory under agents-config/skills/.

Agents (3 Specialized)

  • explore-codebase - Code discovery and analysis
  • Snipper - Fast code modifications
  • websearch - Quick web research

Mods

Installed into ~/.claude/mods/ and enabled through CLAUDE_CODE_PLUGIN_DIRS:

  • aiblueprint - Usage band above the prompt: session context and cost, 5-hour and weekly limits with pace and reset time
  • agent4everything - Links every .agents/skills skill (global and per project) into .claude/skills

Shell Shortcuts

  • cc - Claude Code with permissions skipped
  • ccc - Claude Code with continue mode
  • cx - Codex
  • cxc - Codex continue mode for the current directory

Skills

Install skills directly into ~/.claude/skills/:

# Install all skills
npx skills@latest add melvynx/aiblueprint

# Install a single skill
npx skills add Melvynx/aiblueprint --skill app-icon

# Install multiple skills
npx skills add Melvynx/aiblueprint --skill rules-manager
npx skills add Melvynx/aiblueprint --skill skill-manager

Available skills:

SkillDescription
agents-managerCreate and orchestrate Claude Code agents
apexStructured implementation workflow
app-iconGenerate and prepare app icons
appstore-connectManage App Store Connect workflows
audit-memoriesManually audit and clean project agent documentation
audit-skillsManually audit skill usage, invocation controls, duplicates, and scope
grill-meStress-test a plan with focused questions
hooks-managerCreate and debug Claude Code hooks
environments-managerSet up per-worktree agent environments
rules-managerCreate and maintain agent rule files
skill-managerManage skills and rules across Claude Code, Codex, and Cursor
test-auditGate new tests and prune low-value, implementation-coupled ones
use-styleApply named UI style guides before implementation

🤖 Assistant Pro

Install the latest complete skill bundle from the private assistant-pro-skills repository:

npx aiblueprint-cli@latest assistants pro setup

The command asks for the Assistant Pro access key when needed, installs the versioned all bundle into ~/.agents/skills, creates Claude Code and Codex skill symlinks, adds the shared directory to Hermes, and exposes the same skills to OpenClaw through its native ~/.agents/skills discovery.

💎 Agents Config Pro

Unlock advanced features at mlv.sh/claude-cli

# Setup premium configs (prompts for your token if needed)
npx aiblueprint-cli@latest agents pro setup

# Or activate first if you prefer
npx aiblueprint-cli@latest agents pro activate YOUR_TOKEN

🛠️ Development

# Clone and setup
git clone <repository>
cd aiblueprint-cli
bun install

# Development mode
bun run dev agents setup
bun run dev agents statusline --list

# Run tests
bun run test:run

# Build and test locally
bun run build
bun run test-local

📋 Requirements

  • Node.js 16+ or Bun
  • Claude Code installed
  • Optional: bun, gh CLI
  • Python 3 for assistants pro setup

🤝 Contributing

  1. Fork the repository
  2. Create a feature branch
  3. Make your changes
  4. Run tests: bun run test:run
  5. Submit a pull request

📄 License

MIT License - see LICENSE file for details.

Created by AIBlueprint - mlv.sh/claude-cli

Need help? Open an issue | Documentation

Source 1 files
hooks/register.ts 97 lines
1import type { Register } from 'claude-code'
2
3// `$` as a hook receives it.
4type Engine = Parameters<Parameters<Parameters<Register>[0]>[2]>[0]
5
6type Linked = { root: string; name: string }
7
8const SKILL_FILES = ['SKILL.md', 'skill.md']
9
10const realPath = async ($: Engine, path: string) =>
11  (await $.fs.stat(path, { resolve: true }).catch(() => undefined))?.realPath
12
13const homeDir = async ($: Engine) =>
14  (await $.process.run(['printenv', 'HOME'])).stdout.trim()
15
16// The session's cwd and every ancestor below $HOME, then $HOME itself (global).
17const rootsFor = (cwd: string, home: string) => {
18  const roots: string[] = []
19  let dir = cwd.replace(/\/+$/, '')
20  while (dir.startsWith(`${home}/`)) {
21    roots.push(dir)
22    dir = dir.slice(0, dir.lastIndexOf('/'))
23  }
24  if (!cwd.startsWith(`${home}/`) && cwd !== home) roots.push(cwd)
25  roots.push(home)
26  return roots
27}
28
29const isSkillDir = async ($: Engine, dir: string) => {
30  const stat = await $.fs.stat(dir).catch(() => undefined)
31  if (stat?.kind !== 'dir') return false
32  for (const file of SKILL_FILES) {
33    if (await $.fs.exists(`${dir}/${file}`)) return true
34  }
35  return false
36}
37
38// Links each <root>/.agents/skills/<name> missing from <root>/.claude/skills.
39const syncRoot = async ($: Engine, root: string): Promise<Linked[]> => {
40  const source = `${root}/.agents/skills`
41  const target = `${root}/.claude/skills`
42  if (!(await $.fs.exists(source))) return []
43
44  // .claude/skills already is .agents/skills (a symlinked folder): nothing to map.
45  const sourceReal = await realPath($, source)
46  if (sourceReal !== undefined && sourceReal === (await realPath($, target))) return []
47
48  // Every name already in .claude/skills, dangling links included, case-insensitive.
49  const existing = new Set(
50    (await $.fs.list(target).catch(() => [])).map(entry => entry.name.toLowerCase()),
51  )
52
53  const linked: Linked[] = []
54  for (const entry of await $.fs.list(source)) {
55    const { name } = entry
56    if (name.startsWith('.') || name.startsWith('__')) continue
57    if (existing.has(name.toLowerCase())) continue
58    if (!(await isSkillDir($, `${source}/${name}`))) continue
59
60    if (linked.length === 0) await $.process.run(['mkdir', '-p', target])
61    const ran = await $.process.run(['ln', '-s', `../../.agents/skills/${name}`, `${target}/${name}`])
62    if (ran.exitCode === 0) linked.push({ root, name })
63    else $.ui.log(`agent4everything: ln failed for ${name}: ${ran.stderr.trim()}`, { to: 'debug' })
64  }
65  return linked
66}
67
68const syncAll = async ($: Engine, cwd: string) => {
69  const home = await homeDir($)
70  const linked: Linked[] = []
71  for (const root of rootsFor(cwd, home)) linked.push(...(await syncRoot($, root)))
72  if (linked.length > 0) {
73    $.ui.toast(`agent4everything: linked ${linked.map(l => l.name).join(', ')}`)
74  }
75  return linked
76}
77
78export const register: Register = on => {
79  // Awaited before the first prompt: links exist before the session reads skills again.
80  on('session.start', async ($, e, next) => {
81    await syncAll($, e.cwd).catch(error => $.ui.log(`agent4everything: ${error}`, { to: 'debug' }))
82    return next(e)
83  })
84
85  // Startup, resume, /clear, compact: sync again and have the engine re-read skills.
86  on('classic.SessionStart', async ($, e, next) => {
87    await syncAll($, e.cwd).catch(() => [])
88    const result = await next(e)
89    return { ...result, reloadSkills: true }
90  })
91
92  on('classic.CwdChanged', async ($, e, next) => {
93    await syncAll($, e.new_cwd).catch(() => [])
94    return next(e)
95  })
96}
97