convex-agents

Build persistent AI agents with tool integration, streaming, and workflow orchestration.

Updated Feb 8, 2026

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

Ready to run

npx skills add https://github.com/gursheyss/styx --skill convex-agents-gursheyss

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: convex-agents
Source: https://github.com/gursheyss/styx/tree/main/.agents/skills/convex-agents
Command: npx skills add https://github.com/gursheyss/styx --skill convex-agents-gursheyss

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Creating robust, persistent AI agents is challenging due to fragile state handling, integration of tools, and coordination of long-running tasks.

Core Features & Use Cases

Persistent state and conversation threads across restarts
Tool integration for invoking external functions and services
Streaming responses for real-time user experiences
RAG-based knowledge access and durable workflows for complex tasks
Workflow orchestration to encode multi-step tasks

Quick Start

Create a Convex agent with thread management, tool integration, streaming responses, and workflow orchestration.

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?▼

Persistent AI agents are built using Convex to manage durable state and conversation threads. This approach ensures multi-step tasks and tool execution survive restarts, providing robust agent orchestration.

What is the best way to orchestrate multi-step tasks with external tools in production?▼

Orchestrating multi-step tasks is handled through durable workflows that coordinate external tools and data sources. This enables scalable AI agents to execute complex, long-running tasks reliably in production environments.

How does streaming work with AI agents using Convex?▼

Streaming responses in Convex provide real-time user experiences by continuously delivering agent outputs. This allows production AI agents to stream data directly to clients during tool execution and workflow processing.

Can I use RAG for knowledge access in durable AI workflows?▼

RAG-based knowledge access is fully supported within durable workflows. This integration allows AI agents to query external data sources during persistent threads, ensuring accurate context for complex task coordination.

Does Convex support integrating external functions and services as tools for agents?▼

Convex supports tool integration for invoking external functions and services within AI agents. This enables production-ready agents to execute multi-step tasks by coordinating various external tools alongside persistent state management.