repository·5.x·Indexed 27 days ago
https://github.com/mongodb/laravel-mongodbProvides MongoDB support for Laravel applications by extending Eloquent models and the Query builder. It allows developers to use standard Laravel database APIs with a MongoDB backend, including support for MongoDB as a cache and session driver, BSON type handling via specialized Eloquent models, and integration with the ext-mongodb PHP extension.
Low novelty means the content mostly restates information the model already knows from training data. A score below 3 is a warning sign that the skill may not justify its context window cost.
Common low-novelty patterns:
Advice to fix:
LLM scoring in skill-validator uses the claude-cli provider. This provider relies on a local claude binary and uses your existing CLI authentication rather than API keys.
To use LLM scoring, you must:
claude --version). On macOS, use: curl -fsSL https://claude.ai/install.sh | bash.claude interactively and completing the login process.# Install Claude Code CLI on macOS
curl -fsSL https://claude.ai/install.sh | bash
# Verify installation
claude --versionTo use MongoDB with Eloquent, your models must extend MongoDB\Laravel\Eloquent\Model or use the DocumentModel trait. Use protected $table to define the collection name (note that protected $collection is no longer used). Ensure you cast foreign keys to string in the $casts array to prevent BSON type mismatches during queries.
<?php
namespace App\Models;
use MongoDB\Laravel\Eloquent\Model;
final class Post extends Model
{
protected $connection = 'mongodb';
protected $table = 'posts'; // Use $table, not $collection
protected $fillable = ['title', 'body', 'author_id', 'published_at'];
protected $casts = [
'author_id' => 'string', // Cast FK to string for relationship matching
'published_at' => 'datetime',
];
}When reviewing the content of a SKILL.md file, evaluate it against the following criteria to ensure high quality:
Failing these checks provides non-blocking areas for Subject Matter Experts (SMEs) to address before publishing.
Atlas Auto-embedding is the recommended approach for text fields. Atlas automatically generates and maintains embeddings on insert and update, meaning no PHP code is required to handle vector generation or storage.
To use this approach:
vectorSearchIndex with the autoEmbed type.queryText parameter in an aggregation pipeline, which allows Atlas to auto-embed the search string.<?php
use Illuminate\Database\Migrations\Migration;
use Illuminate\Support\Facades\Schema;
return new class extends Migration
{
public function up(): void
{
Schema::connection('mongodb')->table('products', function ($collection): void {
$collection->vectorSearchIndex([
'fields' => [
[
'type' => 'autoEmbed',
'modality' => 'text',
'path' => 'description', // source text field
'model' => 'voyage-4',
],
['type' => 'filter', 'path' => 'category'],
],
], 'products_vector');
});
}
};Low actionability means an agent cannot follow instructions step-by-step. Common causes include abstract advice, missing intermediate steps, assumptions about unprovided context, or a lack of input/output examples.
Advice to fix:
To use MongoDB, add a mongodb connection entry to your config/database.php file and define the connection details in your .env file.
config/database.php with the mongodb driver configuration.MONGODB_URI and MONGODB_DATABASE in your .env file.// config/database.php
'connections' => [
'mongodb' => [
'driver' => 'mongodb',
'dsn' => env('MONGODB_URI', 'mongodb://localhost:27017'),
'database' => env('MONGODB_DATABASE', 'laravel'),
],
],# .env
MONGODB_URI=mongodb://localhost:27017
MONGODB_DATABASE=laravelTo relate a SQL database (e.g., MySQL) to a MongoDB database, follow these rules:
HybridRelations only to the SQL model. Never add this trait to the MongoDB model._id as a string column (e.g., VARCHAR(24)).string to facilitate direct queries.SQL Model Setup:
Use the MongoDB\Laravel\Eloquent\HybridRelations trait.
MongoDB Model Setup:
Cast the reference field to string and use standard relation methods.
// SQL model (e.g. User in MySQL)
use MongoDB\Laravel\Eloquent\HybridRelations;
final class User extends \Illuminate\Database\Eloquent\Model
{
use HybridRelations; // ONLY on the SQL model
public function posts(): \MongoDB\Laravel\Relations\HasMany
{
return $this->hasMany(\App\Models\Post::class, 'user_id');
}
}
// MongoDB model (e.g. Post)
final class Post extends \MongoDB\Laravel\Eloquent\Model
{
protected $casts = ['user_id' => 'string'];
public function user(): \MongoDB\Laravel\Relations\BelongsTo
{
return $this->belongsTo(User::class, 'user_id');
}
}The methods withCount(), withAvg(), and withSum() are not supported on MongoDB models. To achieve these results, use the raw() method to execute an aggregation pipeline involving $lookup and $size, $avg, or $sum.
// CORRECT
$posts = Post::raw(fn ($c) => $c->aggregate([
['$lookup' => [
'from' => 'comments',
'localField' => '_id',
'foreignField' => 'post_id',
'as' => 'comments',
]],
['$addFields' => ['comments_count' => ['$size' => '$comments']]],
['$project' => ['comments' => 0]],
]));Review reports use a scale of 1-5 for each dimension. The overall score is the mean of all dimensions. Use the following scale to interpret results:
| Score | Meaning |
|---|---|
| 5 | Excellent — genuinely outstanding on this dimension |
| 4 | Good — minor improvements possible but solid |
| 3 | Adequate — functional but has clear room for improvement |
| 2 | Needs work — notable issues that should be addressed |
| 1 | Poor — fundamental problems on this dimension |
Low token efficiency means content is bloated relative to its instructional value. Common causes include redundant explanations, non-helpful boilerplate, verbose phrasing, or unnecessary content.
Advice to fix: