Get started

Convex for AI agents

Convex is a backend-as-a-service platform with a real-time database and serverless functions, letting developers ship an app backend without managing infrastructure. Connect it once through Arc0, and your agent, or Claude, ChatGPT and Cursor, can use it through one MCP endpoint, limited to what each user approved.

Connect ConvexSee the 19 actions ↓
Developer toolsAPI keyMCP + RESTconvex.dev
AUDIT LOG · CONVEXPOLICY: acme-support
09:41:07 · claude · u_8f2read
convex.list_projects
List Projects✓ allowed · 212ms
09:41:08 · claude · u_8f2read
convex.get_deployment
Get Deployment Details✓ allowed · 164ms
09:41:09 · claude · u_8f2write
convex.create_project
Create Project✓ approved · approved by user
09:41:12 · claude · u_8f2destructive
convex.delete_project
Delete project✕ blocked · policy: deny
EVERY CONVEX CALL, ON THE RECORD
01 · USE CASES

What agents do in Convex.

01

Check a deployment's details

Check a deployment's details before deciding whether to update or redeploy it.

02

List a project's deployments

List the deployments under a project to see which environment is active.

03

Confirm before deleting a deployment

Confirm before deleting a deployment, since its data and functions are removed with it.

02 · ACTIONS

19 Convex actions, graded by risk.

Every Convex action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.

read

11

Look things up. Allowed by default.

  • convex.list_projects
    List Projects
  • convex.get_deployment
    Get Deployment Details
  • convex.list_deploy_keys
    List Deploy Keys
  • convex.list_deployments
    List Deployments
  • convex.list_log_streams
    List Log Streams
  • convex.get_project_by_id
    Get Project by ID
  • convex.get_token_details
    Get token details
  • convex.get_project_by_slug
    Get Project by Slug
  • convex.get_query_timestamp
    Get Query Timestamp
  • convex.list_deployment_classes
    List deployment classes
  • convex.list_deployment_regions
    List deployment regions

write

5

Create and change things. Allow, or ask the user first.

  • convex.create_project
    Create Project
  • convex.create_deploy_key
    Create deploy key
  • convex.create_deployment
    Create Deployment
  • convex.update_deployment
    Update Deployment
  • convex.execute_query_batch
    Execute Query Batch

destructive

3

Delete, cancel or archive. Ask first, or deny outright.

  • convex.delete_project
    Delete project
  • convex.delete_deployment
    Delete Deployment
  • convex.delete_custom_domain
    Delete Custom Domain
03 · HOW IT WORKS

Convex in three steps.

  1. 01Your users connect ConvexThey add their Convex api key on Arc0 Connect, under your brand. It goes straight into the vault.
  2. 02You set the rulesReads run, writes like “create Project” can wait for the user, and “delete project” can be denied outright.
  3. 03Any agent can actYour agent calls Convex through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
POLICY.TS
await arc0.policies.set('convex', {
  read: 'allow',
  write: 'ask',        // create_project
  destructive: 'deny',  // delete_project
})

# Claude Code: the same connection, one URL
$ claude mcp add --transport http arc0 \
    https://mcp.arc0.ai/u/u_8f2
04 · AUTH AND DATA

How Convex connects.

Users add their Convex api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.

The same Convex connection serves your agent over MCP and your own backend over REST and the proxy, so a user connects once. How Arc0 handles credentials →

AUTH
API key · Bearer token
CREDENTIALS
Per-tenant encrypted vault
MODEL SEES
Results only, never credentials
AUDIT LOG
Every call, on every plan
05 · WORKS WITH

Use Convex from any agent.

Claude
ChatGPT
Cursor
Codex
VS Code
OpenAI Agents SDK
Claude Agent SDK
Vercel AI SDK
Mastra
LangGraph
06 · RELATED

More developer tools apps.

GitHub896 actionsSupabase129 actions0CodeKit8 actionsAbly25 actionsAbstract3 actionsAddressfinder17 actionsAddresszen2 actionsAlchemy36 actions
ALL DEVELOPER TOOLS APPS →
07 · FAQ

Convex and Arc0, answered.

Q01

Can I use Convex with Claude, ChatGPT or Cursor?

Yes. Connect Convex to Arc0 once, then add your Arc0 MCP URL to Claude, ChatGPT, Cursor, Claude Code or any other remote-MCP client. Each assistant only gets the Convex actions you allow.

Q02

How do users connect Convex?

Users add their Convex api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.

Q03

Which Convex actions can my agent take?

19 in total: 11 read, 5 write and 3 destructive, such as “create Project”. Your policies decide which of them each agent may call.

Q04

Can I stop my agent from deleting things in Convex?

Yes. Actions like “delete project” are graded destructive. Set destructive actions to deny, or to ask so the user approves each one, and blocked calls still show up on the audit log.

Q05

Can my own backend call Convex too?

Yes. The same Convex connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.

Get started

Plug Convex into your agent.

Your users connect Convex once, under your brand. Your agent gets 19 actions behind your policies, with every call on the record.

Connect ConvexBrowse all apps →

Free to build · MCP + REST · Audit log on every plan