/init

description: "Auto-detect project stack and save as memories"
argument-hint: ""

/nemp:init

Intelligently scan your project and auto-save context as memories.

Instructions

You are an intelligent project analyzer. Scan the current project to detect its stack, then save the findings as Nemp memories.

Step 1: Lightweight Scan (Token Optimized)

**IMPORTANT: Use minimal token consumption. Only read essential files. Run ONE combined check.**

echo "=== FILES ===" && ls package.json tsconfig.json next.config.* vite.config.* nuxt.config.* astro.config.* svelte.config.* angular.json requirements.txt pyproject.toml Cargo.toml go.mod Gemfile docker-compose.* Dockerfile .env .env.local .env.example 2>/dev/null && echo "=== LOCKFILE ===" && ls package-lock.json yarn.lock pnpm-lock.yaml bun.lockb 2>/dev/null && echo "=== DIRS ===" && ls -d app pages src/app components/ui 2>/dev/null

**DO NOT** run multiple separate file-existence checks. One command, all checks.

Step 2: Read package.json ONLY (if exists)

**IMPORTANT: This is the ONLY file you need to read for detection. DO NOT read README.md, .env files, or scan directories.**

[ -f "package.json" ] && cat package.json

From package.json, detect everything using the tables in Step 3.

Step 3: Parse package.json

**Detect Framework** from dependencies or devDependencies: | Package | Framework | |---------|-----------| | `next` | Next.js | | `react` (without next) | React (standalone) | | `@remix-run/react` | Remix | | `vue` | Vue.js | | `nuxt` | Nuxt.js | | `astro` | Astro | | `svelte` | Svelte | | `@sveltejs/kit` | SvelteKit | | `express` | Express.js | | `fastify` | Fastify | | `hono` | Hono | | `elysia` | Elysia (Bun) | | `@nestjs/core` | NestJS | | `koa` | Koa |

**Detect Language:**

• `typescript` in devDependencies → TypeScript
• tsconfig.json exists → TypeScript
• Otherwise → JavaScript

**Detect Database/ORM:** | Package | Database/ORM | |---------|--------------| | `prisma` or `@prisma/client` | Prisma ORM | | `drizzle-orm` | Drizzle ORM | | `mongoose` | MongoDB (Mongoose) | | `pg` | PostgreSQL (node-postgres) | | `mysql2` | MySQL | | `better-sqlite3` | SQLite | | `@supabase/supabase-js` | Supabase | | `firebase` | Firebase | | `@planetscale/database` | PlanetScale | | `@neondatabase/serverless` | Neon | | `redis` or `ioredis` | Redis |

**Detect Authentication:** | Package | Auth Solution | |---------|---------------| | `next-auth` or `@auth/nextjs` | NextAuth.js (Auth.js) | | `@clerk/nextjs` | Clerk | | `@supabase/auth-helpers-nextjs` | Supabase Auth | | `lucia` | Lucia Auth | | `passport` | Passport.js | | `@kinde-oss/kinde-auth-nextjs` | Kinde | | `firebase` (with auth imports) | Firebase Auth |

**Detect Styling:** | Package | Styling | |---------|---------| | `tailwindcss` | Tailwind CSS | | `@emotion/react` | Emotion | | `styled-components` | Styled Components | | `sass` | Sass/SCSS | | `@chakra-ui/react` | Chakra UI | | `@mantine/core` | Mantine | | `@radix-ui/react-*` | Radix UI | | `shadcn` (check components/ui) | shadcn/ui |

**Detect Testing:** | Package | Testing | |---------|---------| | `jest` | Jest | | `vitest` | Vitest | | `@testing-library/react` | React Testing Library | | `playwright` | Playwright | | `cypress` | Cypress |

**Detect State Management:** | Package | State | |---------|-------| | `zustand` | Zustand | | `@reduxjs/toolkit` | Redux Toolkit | | `jotai` | Jotai | | `recoil` | Recoil | | `@tanstack/react-query` | TanStack Query | | `swr` | SWR |

Step 4: Display Summary

Show a compact summary: 🔍 NEMP PROJECT SCAN Project: [name from package.json] Framework: [detected] Language: [TypeScript/JavaScript] Database: [detected or "none detected"] Auth: [detected or "none detected"] Styling: [detected or "none detected"] Testing: [detected or "none detected"] Package Manager: [npm/yarn/pnpm/bun from lockfile in Step 1]

Step 5: Save ALL Memories in ONE Write

**CRITICAL: Do NOT call /nemp:save individually for each memory. Build the full JSON object and write once.**

Only include keys for things actually detected. Compress all values to under 100 characters.

{
  "stack": {
    "value": "Next.js 14 + TypeScript + Prisma + PostgreSQL + Tailwind",
    "created": "<ISO-8601>",
    "updated": "<ISO-8601>",
    "agent_id": "nemp-init",
    "tags": ["auto-detected"]
  },
  "framework": {
    "value": "Next.js 14, App Router",
    "created": "<ISO-8601>",
    "updated": "<ISO-8601>",
    "agent_id": "nemp-init",
    "tags": ["auto-detected"]
  },
  "database": {
    "value": "PostgreSQL via Prisma",
    "created": "<ISO-8601>",
    "updated": "<ISO-8601>",
    "agent_id": "nemp-init",
    "tags": ["auto-detected"]
  }
}

Write this to `.nemp/memories.json` in a single operation:

mkdir -p .nemp

Then use the Write tool to save the complete JSON.

If `.nemp/memories.json` already exists, merge: preserve existing keys, only add/update detected keys.

Step 6: Log the Init Operation

echo "[$(date -u +%Y-%m-%dT%H:%M:%SZ)] INIT agent=nemp-init memories_saved=<count>" >> .nemp/access.log

Step 7: Generate MEMORY.md Index

Create or update `.nemp/MEMORY.md`:

# Nemp Memory Index

> Auto-generated. Last updated: [YYYY-MM-DD HH:MM]

## Stored Memories

| Key | Preview | Agent | Updated |
|-----|---------|-------|---------|
| stack | Next.js 14 + TypeScript... | nemp-init | 2026-02-11 |
| database | PostgreSQL via Prisma | nemp-init | 2026-02-11 |
| auth | NextAuth.js with JWT | nemp-init | 2026-02-11 |

## Files

| File | Purpose |
|------|---------|
| `memories.json` | All stored memories (key-value pairs) |
| `access.log` | Read/write audit trail |
| `config.json` | Plugin configuration |
| `MEMORY.md` | This index file |

Step 8: Check Auto-Sync Config (REQUIRED)

After saving all memories, check:

[ -f ".nemp/config.json" ] && cat .nemp/config.json

If `"autoSync": true`, generate and update CLAUDE