Practical overview
When it fits
Use Better Agents when beginning an agent project or when adopting a structured workflow for prompts, behavioral tests, evaluations and assistant-supported development.
Who it is for
- Developers starting projects with agent frameworks such as Agno, Mastra or LangGraph
- Teams using coding assistants to develop and maintain AI agents
- Agent builders who need documented locations for tests, evaluations and versioned prompts
Documented uses
- Initialize a new agent repository with a consistent structure
- Add scenario-based behavioral tests for agent features
- Evaluate RAG, classification or other agent-pipeline components in notebooks
- Manage prompts as versioned YAML files through a registry
- Configure MCP and coding-assistant guidance for an agent project
Requirements & limits
The repository is public and distributed under the MIT license. The documented setup requires Node.js 22+, npm or pnpm, a supported coding assistant, a LangWatch API key and an LLM-provider API key. Anonymous telemetry is enabled by default, with a documented opt-out variable.
- It supports several agent frameworks rather than defining a single agent runtime.
- The generated structure includes Scenario tests, evaluation notebooks, prompt files, .mcp.json and AGENTS.md.
- The documented initialization flow is for creating projects; migration details for arbitrary existing repositories are not provided.
- Using the tool requires both LangWatch and selected LLM-provider API keys.
- The CLI is TypeScript-based and distributed through npm or npx.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Initialize a standardized agent project
Create a new Better Agents project with selected language, framework, coding assistant, LLM provider and API keys.
Before you start
- Node.js 22+
- npm or pnpm
- A supported coding assistant: Claude Code, Cursor, Antigravity or Kilocode
- A LangWatch API key
- An API key for the chosen LLM provider
Steps
Install the package globally with npm, or invoke it with npx.
Run the Better Agents init command for the current or a new directory.
Follow the CLI prompts to select the programming language, agent framework, coding assistant, LLM provider and API keys.
Expected outcome
A project is initialized with the documented Better Agent Structure, including agent code, scenario-test and evaluation directories, prompt files, MCP configuration and AGENTS.md guidance.
Guide limitations
- The source documents initialization and generated artifacts but does not establish successful execution of the resulting agent project.
- The selected framework, provider and assistant determine the project choices presented during initialization.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 3 October 2026
- Original website title
- langwatch/better-agents · Official homepage title · retained in its source language

