pruny 1.0.4 → 1.0.6

package/docs/index.md

@@ -0,0 +1,92 @@
+# Pruny
+
+> Find and remove unused Next.js API routes.
+
+## Summary
+Pruny is a CLI tool that scans Next.js projects (App Router) to identify API routes that are not referenced in the codebase. It supports auto-detection of usage via `fetch`, `axios`, and `useSWR`, and respects `vercel.json` cron jobs.
+
+## Installation
+
+```bash
+npm install -g pruny
+# or run directly
+npx pruny
+```
+
+## Usage
+
+### Scan Mode
+Scans the current directory or specified directory for unused routes.
+```bash
+npx pruny
+npx pruny --dir ./src
+```
+
+### Fix Mode (Deletion)
+Automatically deletes the folders of unused API routes.
+```bash
+npx pruny --fix
+```
+
+### JSON Output
+Outputs the results in JSON format for program consumption.
+```bash
+npx pruny --json
+```
+
+## Configuration
+Pruny looks for `pruny.config.json` in the project root.
+
+```json
+{
+  "dir": "./",
+  "ignore": {
+    "routes": ["/api/webhooks/**", "/api/cron/**"],
+    "folders": ["node_modules", ".next", "dist", ".git"],
+    "files": ["*.test.ts", "*.spec.ts"]
+  },
+  "extensions": [".ts", ".tsx", ".js", ".jsx"]
+}
+```
+
+## Advanced Usage & Limitations
+
+### Dynamic Routes & Custom Clients
+Pruny detects API usage by static analysis of string literals.
+- **Supported**:
+  - `fetch('/api/users')`
+  - `` fetch(`/api/users/${id}`) `` (Template literals with constant prefix)
+  - `axios.get('/api/items')`
+  - `useSWR('/api/data')`
+- **Limitations**:
+  - **Runtime Construction**: Routes constructed purely at runtime without a static `/api/` prefix literal cannot be detected.
+    - ❌ `fetch(myUrlVariable)` -> **False Positive** (Pruny thinks route is unused)
+    - ✅ `fetch('/api/' + myUrlVariable)` -> **Detected**
+  - **Custom Wrappers**: If you wrap `fetch` in a custom `apiClient` without standard method names, use `// pruny-ignore` (future feature) or manually ignore the route in config.
+
+### Monorepo / Multi-App Setup
+For monorepos with multiple Next.js apps (e.g., `apps/web`, `apps/admin`):
+
+1. **Run per package**:
+   ```bash
+   cd apps/web && npx pruny
+   ```
+2. **Run from root with --dir**:
+   ```bash
+   # Scan specific app
+   npx pruny --dir ./apps/web
+   
+   # Scan another app
+   npx pruny --dir ./apps/admin
+   ```
+   *Note: Pruny's config resolution looks for `pruny.config.json` in the specific `dir` being scanned.*
+
+### Vercel Cron Detection
+Pruny automatically parses `vercel.json` to process cron jobs.
+- It looks for `crons: [{ path: "/api/cron/job" }]` array.
+- Any route path found in `vercel.json` is automatically marked as **USED**.
+- This prevents accidental deletion of server-side cron handlers that are never "fetched" by client code.
+
+## Source Code
+- Repository: https://github.com/WebNaresh/pruny
+- Registry: https://www.npmjs.com/package/pruny

package/llms.txt

@@ -49,16 +49,43 @@ Pruny looks for `pruny.config.json` in the project root.
 }
 ```
 
-## How Detection Works
-1. **Discovery**: Finds all `app/api/**/route.{ts,js}` files.
-2. **Scanning**: Greps all source files for usage patterns like:
-   - `fetch('/api/my-route')`
-   - `axios.get('/api/my-route')`
-   - `useSWR('/api/my-route')`
-   - Template literals: `` `/api/${variable}` `` (matches exact prefix)
-3. **Exclusions**:
-   - Routes defined in `vercel.json` crons are automatically marked as used.
-   - Routes matching `ignore.routes` glob patterns are skipped.
+## Advanced Usage & Limitations
+
+### Dynamic Routes & Custom Clients
+Pruny detects API usage by static analysis of string literals.
+- **Supported**:
+  - `fetch('/api/users')`
+  - `` fetch(`/api/users/${id}`) `` (Template literals with constant prefix)
+  - `axios.get('/api/items')`
+  - `useSWR('/api/data')`
+- **Limitations**:
+  - **Runtime Construction**: Routes constructed purely at runtime without a static `/api/` prefix literal cannot be detected.
+    - ❌ `fetch(myUrlVariable)` -> **False Positive** (Pruny thinks route is unused)
+    - ✅ `fetch('/api/' + myUrlVariable)` -> **Detected**
+  - **Custom Wrappers**: If you wrap `fetch` in a custom `apiClient` without standard method names, use `// pruny-ignore` (future feature) or manually ignore the route in config.
+
+### Monorepo / Multi-App Setup
+For monorepos with multiple Next.js apps (e.g., `apps/web`, `apps/admin`):
+
+1. **Run per package**:
+   ```bash
+   cd apps/web && npx pruny
+   ```
+2. **Run from root with --dir**:
+   ```bash
+   # Scan specific app
+   npx pruny --dir ./apps/web
+   
+   # Scan another app
+   npx pruny --dir ./apps/admin
+   ```
+   *Note: Pruny's config resolution looks for `pruny.config.json` in the specific `dir` being scanned.*
+
+### Vercel Cron Detection
+Pruny automatically parses `vercel.json` to process cron jobs.
+- It looks for `crons: [{ path: "/api/cron/job" }]` array.
+- Any route path found in `vercel.json` is automatically marked as **USED**.
+- This prevents accidental deletion of server-side cron handlers that are never "fetched" by client code.
 
 ## Source Code
 - Repository: https://github.com/WebNaresh/pruny

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "pruny",
-  "version": "1.0.4",
+  "version": "1.0.6",
   "description": "Find and remove unused Next.js API routes",
   "type": "module",
   "bin": {