Convex
Backend-as-a-Service platform with real-time database, serverless functions, and built-in authentication
Verdict
Common use cases
- Spin up dev environments during sprint planning
- Execute query batches to debug production data
- Generate deploy keys for CI pipelines
- Inspect deployment details during incident response
- Delete stale projects after feature launches
Integration
- Vendor
- Convex
- Category
- developer-tools
- Auth
- API_KEY
- Tools
- 19
- Composio slug
convex
Tools
- Create deploy key
Tool to create a deploy key for use with the Convex CLI. Use when you need to generate credentials for CLI-based development or deployment workflows. The generated key provides administrative access to the specified deployment.
- Create Deployment
Tool to create a new deployment for a Convex project. Use when you need to create a development, production, or custom deployment. Specify the deployment type and optional configuration like class, reference, and region.
- Create Project
Tool to create a new project on a Convex team, optionally provisioning a dev or prod deployment. Use when you need to initialize a new Convex project in a team.
- Delete Custom Domaindestructive
Tool to remove a custom domain from a Convex deployment. Use when you need to delete a previously configured custom domain.
- Delete Deploymentdestructive
Tool to delete a Convex deployment. Use when you need to permanently remove a deployment and all its data. WARNING: This action will delete all data and files in the deployment and cannot be undone.
- Delete projectdestructive
Deletes a Convex project and all its deployments permanently. Use when you need to permanently remove a project and all associated data. This operation cannot be undone.
- Execute Query Batch
Tool to execute multiple Convex query functions in a single batch request. Use when you need to fetch data from multiple queries efficiently in one API call.
- Get Deployment Details
Tool to retrieve details about a Convex cloud deployment. Use when you need to get information about a specific deployment including its configuration, region, creation time, and status.
- Get Project by ID
Tool to retrieve detailed information about a specific Convex project by its ID. Use when you need to fetch project metadata including name, slug, team association, and creation time.
- Get Project by Slug
Tool to retrieve a Convex project by its slug within a team. Use when you need to fetch project details using human-readable identifiers instead of numeric IDs.
- Get Query Timestamp
Tool to get the latest timestamp for queries from Convex deployment. Use when you need to retrieve the current query timestamp from the Convex API.
- Get token details
Tool to retrieve token details for the authenticated token. Returns the team ID for team tokens or project ID for project tokens. Especially useful after receiving a token from an OAuth flow to identify which team or project it belongs to.
- List Deploy Keys
Tool to list all deploy keys for a specified Convex deployment. Use when you need to view all authentication tokens that can be used to deploy to this deployment.
- List deployment classes
Tool to list available deployment classes for a Convex team. Use when you need to check which deployment classes are available for a specific team.
- List deployment regions
Tool to list available deployment regions for a Convex team. Use when you need to check which regions are available for deploying a team's backend.
- List Deployments
Tool to list all deployments for a Convex project. Use when you need to see all deployments (production, preview, or local) for a specific project.
- List Log Streams
Tool to list all existing log stream configurations in a deployment. Use when you need to view configured log streaming destinations like Datadog, Webhook, Axiom, or Sentry.
- List Projects
Tool to list all projects for a specific Convex team. Use when you need to retrieve all projects associated with a team by team ID.
- Update Deployment
Tool to update properties of an existing Convex deployment. Use when you need to modify deployment settings such as dashboard edit confirmation or deployment reference. Only the fields provided in the request are modified.
Setup
Setup guide
- 11. In Switchy, open the Space where you want to use Convex and click Integrations in the sidebar. 2. Search for Convex and click Connect. 3. You'll be prompted to paste a Convex API key — generate one from your Convex dashboard under Settings > API Keys with admin permissions. 4. Paste the key into Switchy and click Authorize. 5. To test the connection, type '@Convex list my projects' in the Space chat. 6. If the MCP responds with your project list, you're ready. 7. To invoke a tool, @mention Convex in any message and describe what you need — for example, '@Convex create a new dev deployment for project X' or '@Convex execute this query batch'. 8. The MCP will confirm actions that modify data (like deleting a deployment) before proceeding.
What teammates see: by default, memories from Convex are scoped to the Space (PROJECT visibility) - you can mark any memory PRIVATE or share it ORG-wide.
Works well with
Top models
Compatibility data appears once enough Spaces have used this MCP together with a given model.
How Switchy teams use it
Starter prompts
Create Dev Deployment
@Convex create a new development deployment for the 'mobile-app' project with default configurationOpen in a Space →
List All Projects
@Convex show me all projects in our team with their deployment counts and creation datesOpen in a Space →
Execute Query Batch
@Convex execute a batch of queries: fetch user count, recent orders, and active sessions from the production deploymentOpen in a Space →
Generate Deploy Key
@Convex generate a deploy key for the 'api-backend' project so I can configure GitHub ActionsOpen in a Space →
Get Deployment Details
@Convex get full details for the production deployment including region, creation time, and current statusOpen in a Space →
Example outputs
Illustrative - representative of the model's voice and quality, not literal recordings.
@convex list all deployments for our team's project and show me which ones are production vs development
Based on the deployment details retrieved, your team has 3 active deployments: **Production**: `prod-convex-app` (us-east-1, created 45 days ago, status: active) **Development**: `dev-staging` (us-west-2, created 12 days ago, status: active) **Development**: `dev-feature-auth` (us-east-1, created 3 days ago, status: active) The production deployment is running on a standard class instance. Both dev deployments are using the default development configuration. All deployments are healthy with no recent errors.
This example demonstrates read-only inspection of deployment infrastructure. Requires API key with read access to the team's projects. Useful for auditing what environments exist before making changes. The MCP returns structured deployment metadata that the AI can format into a readable summary.
@convex create a new development deployment called 'dev-mobile-redesign' in us-west-2 for testing our mobile UI changes
Successfully created development deployment `dev-mobile-redesign`. **Deployment URL**: https://dev-mobile-redesign.convex.cloud **Region**: us-west-2 **Status**: Provisioning (typically ready in 30-60 seconds) **Deployment ID**: dep_abc123xyz You can now push your schema and functions using the Convex CLI with this deployment URL. The deployment is isolated from your production data and can be safely deleted when testing is complete.
This write action creates real infrastructure in your Convex account. Requires API key with deployment creation permissions. New deployments incur usage charges based on your plan. The MCP confirms the action but doesn't wait for full provisioning—check deployment status separately if you need to verify it's fully ready.
@convex run a batch query to fetch the user count, active sessions, and recent errors from our production deployment, then summarize the health status
Executed batch query against production deployment. Here's the health summary: **Users**: 2,847 total registered users **Active Sessions**: 142 users currently online **Recent Errors** (last 24h): 3 function errors detected - 2× `messages:send` timeout errors (likely transient) - 1× `auth:validateToken` validation failure **Assessment**: System is healthy. Active user ratio (5%) is normal for this time of day. The timeout errors are below the 0.1% threshold and may resolve with retry logic. The auth error appears to be a single malformed request rather than a systemic issue.
This example shows the MCP's batch query capability combined with AI reasoning to interpret results. Requires read access to production data and appropriate query functions already deployed in your Convex backend. The batch execution is more efficient than individual queries but still counts against your function execution limits.
Use-case deep-dives
When Convex MCP handles dev-to-prod promotion for backend teams
A 6-person backend team shipping a real-time collaboration app needs to spin up preview deployments for every feature branch, then promote stable builds to production. The Convex MCP wins here because it exposes deployment lifecycle tools (create, delete, get details) that let the team script environment management directly in Switchy workflows. One engineer writes a prompt that creates a deployment, runs schema migrations, and configures custom domains—then the whole team reuses it. The API key auth means no OAuth dance per developer. The threshold: if your team already has CI/CD pipelines handling this, the MCP adds friction. But if you're manually clicking through dashboards to provision environments, this cuts that toil to a single Switchy command.
Why Convex MCP works for support teams querying production state
A 3-person support team at a SaaS startup needs to answer customer questions by checking live database state—user subscription status, feature flags, recent activity logs. The Convex MCP's Execute Query Batch tool lets support reps run read-only queries against production without writing SQL or bugging engineering. They type natural language in Switchy, the MCP translates it to Convex queries, and returns structured results. The win is speed: no waiting for an engineer to write a one-off script. The boundary: this only works if your Convex schema is already designed with support queries in mind (indexed fields, clear naming). If your data model is a mess, the MCP will surface that mess faster. For teams with clean schemas and frequent support lookups, this pays off in the first week.
When Convex MCP accelerates junior dev onboarding
A startup hiring its first two junior engineers needs them productive on the Convex backend without a week of reading docs. The Convex MCP shortens that ramp because it surfaces the 19 available tools as discoverable actions in Switchy—new devs see what's possible (create projects, manage deploy keys, batch queries) without memorizing CLI flags. A senior engineer writes a few example prompts ("set up my local dev deployment", "check prod database schema") and juniors copy-paste-modify them. The API key setup is one Slack message, not a multi-step OAuth flow. The limit: if your team is already fluent in Convex CLI, the MCP is redundant overhead. But for teams onboarding non-experts or mixing technical and semi-technical roles, this flattens the learning curve to hours instead of days.
Frequently asked
What does the Convex MCP let me do in Switchy?
It lets your team manage Convex projects, deployments, and domains without leaving the chat. You can create new projects, spin up dev or prod deployments, execute query batches, and generate deploy keys for CLI workflows. It's useful when you're coordinating backend changes across multiple deployments or need to quickly check deployment status during a conversation.
Do I need admin access to connect Convex MCP?
Yes. The MCP uses an API key that grants administrative access to your Convex team, including the ability to create and delete projects and deployments. Only connect it if you're comfortable giving the AI full control over your Convex infrastructure. You generate the key in your Convex dashboard under team settings.
Can the Convex MCP write or modify my database records?
No. It can execute query batches to read data, but it doesn't have tools to run mutations or actions. If you need to write data, you'll still use the Convex dashboard, CLI, or your app's mutation functions. The MCP is for infrastructure management and read-only data inspection, not for changing records.
Why use this instead of the Convex dashboard or CLI?
Use the MCP when you're already in a Switchy conversation and need to check deployment details, create a new environment, or batch-query data without context-switching. The dashboard is better for visual schema browsing and the CLI is better for scripted deploys. The MCP shines for ad-hoc ops tasks during team discussions.
Who on my team should connect the Convex MCP?
Whoever owns your Convex infrastructure — typically a backend lead or DevOps engineer. Because the API key can delete deployments and projects, don't share the connection with people who shouldn't have that level of access. Each Switchy workspace member with the connection can use all 19 tools, so treat it like you would admin credentials.