convex-agents

Develop AI agents with Convex for thread management, tool integration, and RAG patterns.

Updated Mar 7, 2026

Search Tags:#llm #rag #ai agents #tools #workflows #convex

Ready to run

npx skills add https://github.com/pratikpakhale/convex-next-starter --skill convex-agents-pratikpakhale

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: convex-agents
Source: https://github.com/pratikpakhale/convex-next-starter/tree/main/.agent/skills/convex/modules/convex-agents
Command: npx skills add https://github.com/pratikpakhale/convex-next-starter --skill convex-agents-pratikpakhale

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires @convex-dev/agent, ai, openai, and includes references (resource) components.

What problem does it solve?

This Skill provides a robust framework for building persistent, stateful AI agents that can manage conversations, integrate tools, stream responses, and orchestrate complex workflows using the Convex platform.

Core Features & Use Cases

Thread Management: Create, list, and retrieve conversation threads and messages.
Tool Integration: Define and utilize custom tools for agents to interact with external functionalities or data.
RAG Patterns: Implement Retrieval Augmented Generation for knowledge retrieval and context-aware responses.
Workflow Orchestration: Design multi-step agentic workflows for complex tasks.
Use Case: Develop a customer support agent that can access a knowledge base, create support tickets, and provide real-time assistance to users, all within a persistent chat interface.

Quick Start

Use the convex-agents skill to create a new conversation thread for the current user.

Frequently Asked Questions about convex-agents

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

FAQPage Schema

How do I build stateful AI agents with persistent conversation threads?▼

To build stateful AI agents with persistent conversation threads, use the Convex Agent component to manage thread lifecycles, store messages, and maintain durable state across interactions via the real-time backend.

Can I integrate custom tools into AI agent workflows using Convex?▼

Yes, you can integrate custom tools into AI agent workflows using Convex. The framework allows you to define and utilize external tools, enabling agents to interact with external functionalities and orchestrate complex multi-step tasks.

Does the convex-agents skill support Retrieval Augmented Generation?▼

Yes, the convex-agents skill supports Retrieval Augmented Generation (RAG) patterns. It enables developers to implement knowledge retrieval and context-aware responses directly within the AI agent's workflow.

What's the best way to stream LLM responses for a real-time chat application?▼

The best way to stream LLM responses for real-time chat applications is using the Convex Agent component, which provides built-in streaming response capabilities and leverages Convex's real-time backend for immediate UI updates.

Do I need the OpenAI SDK to orchestrate multi-step agentic workflows?▼

You need the OpenAI SDK and the AI package as dependencies to orchestrate multi-step agentic workflows. These packages provide the core LLM interfaces required to execute complex tasks and durable workflows within Convex.