convex-agents

Build persistent, stateful AI agents with tool hooks and streaming in Convex.

Updated Feb 1, 2026
One-click install
npx skills add https://github.com/EricJamesCrow/smartpockets --skill convex-agents-ericjamescrow
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: convex-agents
Source: https://github.com/EricJamesCrow/smartpockets/tree/main/.claude/skills/convex-agents
Command: npx skills add https://github.com/EricJamesCrow/smartpockets --skill convex-agents-ericjamescrow

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Convex Agents enables developers to build persistent, stateful AI agents that can manage conversations, invoke tools, stream outputs, and orchestrate complex workflows within Convex.

Core Features & Use Cases

  • Persistent state across restarts for agents and threads.
  • Real-time streaming responses with integrated tool calls.
  • RAG capabilities for retrieval-augmented generation and knowledge workflows.
  • Workflow orchestration for multi-step, durable tasks across domains.

Quick Start

Install and initialize a Convex Agent with thread management, tool integration, and streaming chat to begin building a persistent AI agent workflow.

Frequently Asked Questions about convex-agents

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build persistent AI agents that maintain state across restarts?

Build persistent, stateful AI agents within Convex to manage threads, run tools, and stream responses. Agents maintain persistent state across restarts for long-running tasks and multi-turn conversations.

Can I stream real-time AI responses and invoke tools in a single workflow?

Yes, real-time streaming responses with integrated tool calls are supported. Convex agents orchestrate complex workflows, enabling tool invocation and knowledge retrieval in a single integrated flow.

How do I add RAG capabilities to an AI agent workflow?

Add RAG capabilities through Convex components for retrieval-augmented generation and knowledge workflows. This enables persistent agents to retrieve knowledge during multi-step, durable tasks.

Does Convex support durable workflow orchestration for multi-step agent tasks?

Convex supports workflow orchestration for multi-step, durable tasks across domains. It provides durable workflows, tool hooks, and persistent state to handle long-running agent tasks effectively.

What is the best way to manage multi-turn conversations with stateful agents?

The best way to manage multi-turn conversations is using Convex agents with thread management. This approach provides persistent state, real-time streaming, and integrated tool hooks for durable conversational workflows.

Are there limitations when running long-running agent tasks in Convex?

Convex agents are designed for long-running agent tasks with durable workflows and persistent state. Limitations depend on your specific orchestration complexity and the scale of your knowledge retrieval workflows.