JETIC CLI · MCP INTEGRATION
Connect Jetic to Claude Code, Cursor, Codex, and more.
Bring code-aware API testing into your AI coding workflow. Connect Jetic to your preferred AI agent through MCP, give it context about your Node.js API, and work with endpoints, schemas, and simulations without manually navigating every step.
Explore Jetic Agent →
Codex
Connect Codex to Jetic to inspect API behavior, work with available testing tools, and investigate simulation results. Get Codex →
Claude Code
Give Claude access to Jetic's MCP tools to work with your API model and execute supported testing workflows. Get Claude Code →
Cursor
Bring Jetic's code-aware API testing context into your Cursor workflow through MCP. Get Cursor →
Google Antigravity
Connect a compatible Antigravity setup to Jetic's available MCP tools and API testing workflows. Explore Antigravity →
Connect your AI agent to Jetic in three steps
Give your AI coding agent access to Jetic's API testing tools and start working with your project in natural language.
Node 22 · your project
/Users/paul
> npx jetic init
> npx jetic mcp
1. Set up Jetic CLI
Install or initialize Jetic in your Node.js project, then start the MCP integration. You may not need to initialize the project again if it is already configured.
● Configure your MCP client
Add the Jetic MCP server to the agent you use, using the command and settings it requires.
> npx jetic mcp
2. Connect your AI agent
Add Jetic's MCP server to your supported AI coding environment. Configure the server using the command and settings required by that client. Verify client-specific configuration before relying on a particular setup.
● Connected to your Node.js API project
What would you like to work on?
> Inspect my API and identify the endpoints that need behavior tests.
3. Run your first API testing workflow
Ask your agent to inspect an endpoint, generate a simulation, or investigate a failing API request using the Jetic tools available to it. The agent's actual actions depend on the tools registered with the MCP server and the permissions you grant.
What AI agents can do with Jetic
Give your AI coding agent the context and tools to work with your API testing workflow, from understanding endpoints to investigating simulation results.
Endpoints
Explore your API model and understand the endpoints available for testing.
Schemas & dependencies
Use available schema and dependency context to understand API inputs, relationships, and expected behavior.
Simulations
Create and run supported API simulations through natural-language instructions and available Jetic tools.
Behavior validation
Validate API responses and expected behavior against the checks supported by your testing workflows.
Failure investigation
Investigate failed requests and simulation results with help from your AI coding agent.
CLI & MCP workflows
Bring supported Jetic operations directly into compatible AI coding environments.
Frequently asked questions
- Do I need to configure an MCP server manually?
- Jetic exposes its available API testing tools through MCP. You need to configure your AI coding agent to connect to Jetic using the supported CLI command and the configuration required by your client.
- Which AI coding agents does Jetic support?
- Jetic is designed to work with AI coding environments that support MCP, including compatible configurations of Claude Code, Cursor, and Codex. Exact setup steps and available capabilities depend on each client's MCP support.
- What can my AI agent do with Jetic?
- Your agent can use Jetic's exposed tools to work with supported API testing workflows. Depending on the configured tools, this can include inspecting API context, working with endpoints, creating simulations, and investigating results.
- Does Jetic understand my codebase?
- Jetic is designed around code-aware API testing. Its project model provides API context that can help compatible agents work with your Node.js application's endpoints and behavior. The available context depends on your project's configuration and indexing.
- Can the agent run API simulations for me?
- Yes, when the relevant simulation tools are exposed and configured. The agent can use natural-language instructions to initiate supported workflows instead of requiring you to navigate every step manually. Execution permissions and environment configuration still apply.
- Does Jetic use AI tokens every time I run a simulation?
- Jetic's core value proposition is deterministic API workflows without requiring AI inference for every run. AI usage depends on whether you invoke an AI agent or an AI-powered feature; it is separate from deterministic simulation execution.
- Is it safe to connect an AI agent to my API?
- Treat the connection like any other development tool with access to your project. Review the permissions granted to the agent, protect credentials, use appropriate test environments, and verify any state-changing operations before execution. Actual access depends on the MCP tools and permissions configured.
- What happens if a simulation fails?
- Use the simulation results and available API context to investigate the failure. A connected agent can help analyze the output and suggest a fix, while you retain control over code changes and verification.
- Do I need Jetic's web dashboard to use the CLI?
- The CLI and MCP integration are intended to bring Jetic workflows into your development environment. The dashboard can provide a visual interface for inspecting your API model, simulations, and reports, depending on which features you use.