Skip to content
Get packs

Convex AI agent backend

Build an in-app AI agent with durable threads, message history, tools and RAG on Convex

Source Download skill
Repository
openclaw/clawhub
Stars
9,546
Licence
MIT
Files
1 file

What it does

This skill teaches an AI agent how to add a chatbot-style AI agent backend to an app built on Convex, a hosted backend platform. It guides the setup of the @convex-dev/agent package so conversations are saved, tools can be called, and documents can be searched by meaning (retrieval-augmented generation). Model calls go through the Convex AI Gateway by default, so the app does not need its own model provider API key in most cases.

Great for

  • Developers building an AI assistant or chat feature inside a Convex application
  • Teams that want saved conversation history and reactive updates for AI chats
  • Projects that need document search (RAG) over their own content in Convex

Skip it if

  • Apps not using Convex, since the setup steps assume Convex functions, config and deployments (the skill does not help with other backends)
  • Free-plan, self-hosted or local Convex setups, where the gateway is unavailable and a provider key must be stored in Convex env, adding key-handling responsibility
  • Projects that must avoid any new dependency or server-side code execution, because the setup installs packages and adds backend functions

Try asking

  • “Add an AI agent with chat history to my Convex app using @convex-dev/agent”
  • “Set up RAG so my assistant can search the documents I embed in Convex”
  • “Define an agent with a tool that looks up order status and streams replies to the UI”
Step-by-step guide: how to use Convex AI agent backendInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the MIT licence and an attribution note.

    convex-agent.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip convex-agent.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip convex-agent.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Keep the default gateway setup so no model API key has to be created, stored or rotated.
  • If you need embeddings, store that single provider key only in Convex env through the env capability and never in client code.
  • Check that your deployment is on a Convex Cloud paid plan before relying on the gateway, or plan a fallback.

What's inside

  • SKILL.md

Redistributed unmodified from openclaw/clawhub under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.