MongoDB Agent Skills
Official MongoDB skills for natural-language queries, schema design, indexing, Atlas Search and driver connection setup
- 198 ★ on the repository
- Last commit
- Apache-2.0
Extended access
Why this level
- The plugin connects an MCP server that reads cluster data and, with Atlas API credentials, manages stream processing.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/mongodb-agent-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The plugin connects an MCP server that reads cluster data and, with Atlas API credentials, manages stream processing.
This is third-party code. Review the repository files before installing.
What it does
MongoDB Agent Skills is a set of eight skills released by MongoDB itself for Claude Code, Cursor, Codex, GitHub Copilot and other agents. The skills teach the agent practices used by the database's developers and work together with the MongoDB MCP server. `mongodb-natural-language-querying` turns a plain-language request into a `find` query or an aggregation pipeline after studying the collection schema and sample documents. `mongodb-schema-design` walks through modeling patterns and mistakes: embedding versus references, unbounded arrays, migration from SQL. `mongodb-query-optimizer` finds why queries are slow and suggests indexes. `mongodb-search-and-ai` explains full-text, vector and hybrid search in Atlas. Three more skills: `mongodb-connection` tunes client pools and timeouts for different drivers, `mongodb-atlas-stream-processing` manages Atlas stream processing, and `mongodb-qe-size-estimation` estimates the storage and memory Queryable Encryption will take. `mongodb-mcp-setup` guides you through configuring the MCP server itself. There are two ways to install. The `mongodb-atlas` plugin installs from agent marketplaces and connects the hosted Atlas MCP server over OAuth right away. For a self-managed cluster there is the `mongodb` plugin, which runs the MCP server locally.
Who it is for. Developers of MongoDB applications who write queries, design schemas and investigate slow operations together with an agent.
Good fit when
- You need to pick documents by conditions described in plain words and do not want to recall aggregation syntax.
- An application moves off SQL and you must decide what to embed in a document and what to reference.
- A collection query takes seconds and the agent has to find the index that fixes it.
- A serverless function runs out of database connections and the driver pool needs proper settings.
- Semantic search is being added to existing data and you choose between Atlas Search, vector and hybrid search.
Not a fit when
- The database is not MongoDB: the skills are tied to its queries, indexes and Atlas services.
- The agent has to change data: the query-generation skill produces read-only queries only.
- You want optimization without any performance request: the index skill triggers only on questions about speed and indexes.
- There is no Atlas API access: the stream processing skill requires an MCP server with Atlas API credentials.
Example request
In the orders collection find orders from the last week over 5000 and group them by city, starting with a look at the schema.Limitations
The skills work best together with the MongoDB MCP server. Without it the optimizer suggests indexes from query shape only, and the stream processing skill needs Atlas API credentials. The `mongodb-atlas` plugin signs in to Atlas over OAuth in a browser. For Community and Enterprise Advanced use the `mongodb` plugin, which runs the server locally, and its settings are described in `README.community.md`. Installing the skills alone does not connect an MCP server: add it with `npx "mongodb-mcp-server@latest" setup`. Some skills, such as the one on field encryption, target narrow scenarios.
How to disable. Disable or remove the `mongodb-atlas` plugin through the agent's plugin menu. Skills installed with `npx skills add` are deleted from the agent's skills directory. Revoke the OAuth access to Atlas in the MongoDB account settings.
FAQ
How do I install MongoDB Agent Skills in Claude Code?
Run `/plugin install mongodb-atlas`, follow the prompts and run `/reload-plugins`. Without plugins use `npx skills add mongodb/agent-skills`.
How do the mongodb-atlas and mongodb plugins differ?
`mongodb-atlas` connects the hosted Atlas MCP server through OAuth. `mongodb` runs the MCP server locally for Community and Enterprise Advanced and works with your own deployment.
Does installing the skills directly connect an MCP server?
No. Installing with `npx skills add` or copying the `skills/` folder gives only the skills. The server is added by `npx "mongodb-mcp-server@latest" setup`.
Which agents have a plugin?
Claude, Cursor, Codex, GitHub Copilot (CLI and VS Code) and Grok. Cursor installs it with `/add-plugin mongodb-atlas`, Copilot CLI with `copilot plugin install mongodb-atlas`.
Can the query skill change data?
No. `mongodb-natural-language-querying` produces read-only `find` queries and aggregation pipelines. Atlas Search is not handled by this skill.
Discussion
No questions or reviews yet.
Related
SkillsRedis's official 15 skills teach an agent data structures, search, clustering, security and agent memory, plus a docs MCP server.
PlanetScale Database Skills
planetscale-database-skills
Skills for agents working with databases: one subfolder per technology. Made by PlanetScale, with a separate plugin for Cursor
Official Firebase plugin for agents: 14 skills on Firestore, Auth, Hosting, Crashlytics, Remote Config and security rules
The official Payload CMS skill shipped inside the package: collections, fields, hooks, access control, queries and database adapters
The official Tencent CloudBase kit for coding agents: a plugin installs the stack, skills set the rules, MCP runs the backend
Thirty-two Tencent CloudBase skills: serverless functions, auth, AI models, storage and backend deployment in the cloud.
README in short
The README is about the `mongodb-atlas` plugin and splits into installation for six agents, OAuth sign-in and installing the skills alone without a plugin. For a self-managed deployment the README points to `README.community.md`.
README badge
Original SKILL.md
---
name: mongodb-natural-language-querying
description: Generate read-only MongoDB queries (find) or aggregation pipelines using natural language, with collection schema context and sample documents. Use this skill whenever the user asks to write, create, or generate MongoDB queries, wants to filter/query/aggregate data in MongoDB, asks "how do I query...", needs help with query syntax, or discusses finding/filtering/grouping MongoDB documents. Also use for translating SQL-like requests to MongoDB syntax. Does NOT handle Atlas Search ($search operator), vector/semantic search ($vectorSearch operator), fuzzy matching, autocomplete indexes, or relevance scoring - use search-and-ai for those. Does NOT analyze or optimize existing queries - use mongodb-query-optimizer for that. Does NOT handle aggregation pipelines that involve write operations. Requires MongoDB MCP server.
license: Apache-2.0
metadata:
version: "1.0.0"
allowed-tools: mcp__mongodb__*
---
# MongoDB Natural Language Querying
You are an expert MongoDB read-only query and aggregation pipeline generator.
## Query Generation Process
### 1. Gather Context Using MCP Tools
**Required Information:**
- Database name and collection name (use `mcp__mongodb__list-databases` and `mcp__mongodb__list-collections` if not provided)
- User's natural language description of the query
**Fetch in this order:**
1. **Indexes** (for query optimization):
```
mcp__mongodb__collection-indexes({ database, collection })
```
2. **Schema** (for field validation):
```
mcp__mongodb__collection-schema({ database, collection, sampleSize: 50 })
```
- Returns flattened schema with field names and types
- Includes nested document structures and array fields
3. **Sample documents** (for understanding data patterns):
```
mcp__mongodb__find({ database, collection, limit: 4 })
```
- Shows actual data values and formats