eve
vercel/eve
Filesystem-first framework for building durable backend AI agents with composable instructions, skills, tools, and connections.
What is eve?
eve is a framework for creating production-grade backend AI agents where all configuration—instructions, skills, tools, connections, channels, subagents, schedules, and evals—lives as files on disk. Use it when building, editing, or debugging agent projects that need durability and composability.
- Define agent behavior through filesystem-based instructions and configuration files
- Compose reusable skills and tools into agent capabilities
- Manage agent connections, channels, and communication patterns
- Create subagents and orchestrate multi-agent workflows
- Schedule agent tasks and runs
- Define and run agent evaluations (evals)
How to install eve
npx skills add null --skill eve- Node.js installed
- npm or package manager for installing eve
- Read bundled documentation in node_modules/eve/docs/ after installation
How to use eve
- 1.Install eve with npm install eve or scaffold a new agent with npx eve init <agent-name>
- 2.Read node_modules/eve/docs/README.md for the complete index and recommended reading order
- 3.Create your agent directory structure with instructions, skills, tools, and connections as files
- 4.Define agent behavior through filesystem configuration matching the eve schema
- 5.Compile and run your agent using eve's build and execution tools
- 6.Iterate on agent capabilities by editing files and recompiling
Use cases
- Building production backend agents that persist state and configuration on disk
- Creating multi-agent systems with subagents and inter-agent communication
- Developing agents with scheduled tasks or recurring operations
- Testing and evaluating agent behavior with evals
- Composing complex agent capabilities from modular skills and tools
- Backend engineers building AI-powered services
- Teams creating production-grade autonomous agents
- Developers building multi-agent systems
- AI engineers evaluating and iterating on agent behavior
eve FAQ
Complete documentation ships inside the eve package at node_modules/eve/docs/. Always read the bundled docs that match your installed version exactly, starting with node_modules/eve/docs/README.md.
Use npx eve init <agent-name> to scaffold a new agent project, or manually create an agent directory with the required configuration files. Then read the bundled documentation to understand the structure.
eve agents are defined through filesystem files including: instructions, skills, tools, connections, channels, subagents, schedules, and evals. Each component is a separate file that eve compiles together.
Yes, eve is designed for durable backend AI agents with persistent state and configuration. It's built for production-grade autonomous systems.
Yes, eve supports subagents and multi-agent orchestration, allowing you to build complex systems where agents can communicate and coordinate.
Full instructions (SKILL.md)
Source of truth, from vercel/eve.
name: eve description: Build durable backend AI agents with the eve framework. Use when creating, editing, or debugging an eve project — agent instructions, skills, tools, connections, channels, sandboxes, subagents, schedules, or evals.
eve
eve is a filesystem-first framework for durable backend AI agents. An agent is a directory on disk — instructions, skills, tools, connections, channels, subagents, and schedules are all files — and eve compiles and runs it.
Source of truth
The complete documentation ships inside the eve package. Do not rely on this
skill for guidance — always read the bundled docs, which match the installed
version exactly:
node_modules/eve/docs/
Start with node_modules/eve/docs/README.md. It contains the full
index and recommended reading order. Before writing any eve code, read the
relevant guide there first.
If eve is not installed yet, install it (npm install eve) or scaffold a new
agent with npx eve init <agent-name>, then read the bundled docs.
Related skills
More from vercel/eve and the wider catalog.
flags-sdk
Feature flags and A/B testing SDK for Next.js and SvelteKit with Vercel Flags provider and multi-adapter support.
next-cache-components-adoption
Agent skill from vercel/next.js.
next-cache-components-optimizer
Agent skill from vercel/next.js.
next-dev-loop
Agent skill from vercel/next.js.
next-best-practices
Next.js best practices - file conventions, RSC boundaries, data patterns, async APIs, metadata, error handling, route handlers, image/font optimization, bundling
next-cache-components
Next.js 16 Cache Components - PPR, use cache directive, cacheLife, cacheTag, updateTag