Connect Convex to Claude, ChatGPT, Cursor, Claude Code, Codex, Gemini, GitHub Copilot and any other MCP client, for your whole team. Metorial adds security, governance, and tracing, and gives everyone one Magic MCP URL to connect.
Ask your agent in plain language. Each of these maps to a real tool on the server.
Use the run mutation tool in Convex
Use the generate upload url tool in Convex
Use the get document changes tool in Convex
Use the run query tool in Convex
Use the list documents tool in Convex
Use the run action tool in Convex
Supported Tools
run_mutation
Run Mutation
Execute a Convex mutation function over HTTP. Mutations are write operations that can read and modify data in the Convex database with ACID transaction guarantees.
Specify the function path (e.g. `messages:send` or `users:create`) and optional arguments.
Returns the mutation result.
generate_upload_url
Generate Upload URL
Generate a temporary upload URL for uploading a file to Convex file storage.
The returned URL can be used to POST file content directly. After uploading, a storage ID is returned that can be referenced in documents.
Requires a deployed mutation that returns ctx.storage.generateUploadUrl().
get_document_deltas
Get Document Changes
Retrieve incremental document changes (inserts, updates, deletes) from a Convex deployment using the Streaming Export API.
Returns a cursor-based change feed of document-level modifications since a given cursor position.
Useful for syncing external systems, building audit trails, or tracking data changes over time.
Requires deployment administration permission and Streaming Export availability on your plan.
run_query
Run Query
Execute a Convex query function over HTTP. Queries are read-only functions that fetch data from the Convex database.
Specify the function path (e.g. `messages:list` or `users:getById`) and optional arguments.
Returns the query result as JSON.
list_documents
List Documents
List documents from a Convex table using the Streaming Export API snapshot endpoint.
Retrieves a paginated snapshot of documents from a specific table or across all tables.
Useful for browsing data, exporting records, or inspecting table contents.
Requires deployment administration permission and Streaming Export availability on your plan.
run_action
Run Action
Execute a Convex action function over HTTP. Actions can call external APIs, use third-party libraries, and perform side effects that are not possible in queries or mutations.
Specify the function path (e.g. `ai:generateEmbedding` or `email:send`) and optional arguments.
Returns the action result.
manage_environment_variables
Manage Environment Variables
View or update environment variables for a Convex deployment.
When no changes are provided, lists all current environment variables.
When changes are provided, updates the specified environment variables; omit a change's value to delete that variable.
Listing includes secret values. Requires deployment environment-variable permission.
get_deployment_info
Get Deployment Info
Identify the configured Convex deployment, including its project, team and deployment type when hosted in Convex Cloud.
download_file
Download File
Download a Convex storage file using a deployed query that returns ctx.storage.getUrl(storageId). The query controls access to the file.
Enable Convex in Metorial. In your Metorial portal, enable the Convex MCP server and sign in with the account that owns the data. Access stays governed by your team policies.
Add a custom connector. Open Customize, then Connectors. Select Add, then Custom, then Web.
Paste your Metorial MCP URL. Name it Metorial, paste your Magic MCP URL (<your-magic-mcp-url>) and continue.
Sign in. Select Add and sign in to load Convex tools.
Enable Convex in Metorial. In your Metorial portal, enable the Convex MCP server and sign in with the account that owns the data. Access stays governed by your team policies.
Add a custom MCP server. On ChatGPT web, open Plugins, select +, then Add custom MCP server.
Paste your Metorial MCP URL. Name it Metorial, paste your Magic MCP URL (<your-magic-mcp-url>), choose OAuth and accept the risk notice.
Create and install the plugin. Select Create as a plugin, install it and sign in to load Convex tools.
Enable Convex in Metorial. In your Metorial portal, enable the Convex MCP server and sign in with the account that owns the data. Access stays governed by your team policies.
Add Metorial to your MCP config. Use ~/.cursor/mcp.json for every project, or .cursor/mcp.json for one.
Sign in. Open Customize in the sidebar, find Metorial and sign in to load Convex tools.
{
"mcpServers":{
"metorial":{
"url":"<your-magic-mcp-url>"
}
}
}
Enable Convex in Metorial. In your Metorial portal, enable the Convex MCP server and sign in with the account that owns the data. Access stays governed by your team policies.
Add the server over HTTP. Add --scope user to use it in every project.
Sign in. Run /mcp, select Metorial and sign in to load Convex tools.
claude mcp add --transport http metorial <your-magic-mcp-url>
Enable Convex in Metorial. In your Metorial portal, enable the Convex MCP server and sign in with the account that owns the data. Access stays governed by your team policies.
Add the server and sign in. The Codex CLI, IDE extension and ChatGPT desktop app share this config.
Check the connection. Type /mcp in Codex to see Metorial and its Convex tools.
codex mcp add metorial --url <your-magic-mcp-url>
codex mcp login metorial
Run Convex through the Metorial gateway
One governed layer for every app your agents reach, not a separate connector per tool.
Give Convex and every other app to your whole team through one MCP gateway, behind your SSO.
Scope which teams and agents can use Convex tools with role-based access control, and trace every call.
Prefer to run it yourself? Metorial is open source and self-hostable.
Invoke serverless queries, mutations, and actions on a reactive document-relational database. Read, insert, update, and delete documents across tables with ACID transactions. Upload and manage files in built-in storage. Schedule functions for future or recurring execution (cron jobs). Perform full-text search with BM25 relevance scoring and vector similarity search for AI/RAG use cases. Stream export document-level changes (inserts, updates, deletes) for incremental data sync to external systems. Stream import data from external sources into Convex tables. Manage projects, deployments, deploy keys, and environment variables via the Management API. Define custom HTTP endpoints for inbound webhooks and REST-style APIs.
Frequently asked questions
Common questions about connecting Convex to AI agents with Metorial.
How many tools does the Convex MCP server have?
The Convex MCP server exposes 9 tools, including Run Mutation, Generate Upload URL, Get Document Changes. Each one is governed and traced through Metorial.
Does the Convex MCP server support triggers?
Convex exposes tools your agents call on demand; it does not publish event triggers yet. Everything still runs through one governed MCP URL.
Which AI clients can use the Convex MCP server?
Any MCP client, including VS Code, Windsurf, Zed, OpenCode and Grok. Each one reaches the Convex MCP server through the same Metorial URL.
How do I give only some teams access to Convex?
Use role-based access control to decide which teams, users, and agents can use Convex and which of its tools they can call.
Can I self-host the Convex MCP server?
Yes. Metorial is open source, so you can run the Convex MCP server yourself or use the hosted gateway.
Connect Convex to production AI agents
See how Metorial gives your team governed access to Convex, with tracing and security controls built in.