Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Invoker MCP Server scan my API collection at ./api-tests and run the user registration request"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Let AI agents create, run, and manage HTTP requests in .ivk format.
An MCP (Model Context Protocol) server that wraps ivkjs to let AI tools interact with the Invoker API client.
npm install -g invoker-mcpRelated MCP server: Swagger/Postman MCP Server
Add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"invoker": {
"command": "npx",
"args": ["invoker-mcp"]
}
}
}claude mcp add invoker -- npx invoker-mcpTool | Description |
| Scan a folder for .ivk and .md files |
| Read and parse an .ivk file |
| Create a new .ivk request file |
| Execute a request (from file or inline) |
| List all environments with variables |
| Set an environment variable |
| Switch the active environment |
| Create a new environment |
| Build a static documentation site |
"Scan my API collection at ./invoker and send the login request"
The AI agent would:
Call invoker_list_collection with folder_path: "./invoker"
Call invoker_send_request with file_path: "./invoker/auth/login.ivk"
Return the response to you
MIT
Create a new environment with optional initial variables.
Inputs: name (required), variables.
Build a static documentation site from an Invoker collection folder. Produces a deployable dist/ with markdown docs and inline runnable .ivk request widgets. Requires the invoker-app to be installed.
Inputs: collection_path (required), output_path, title.
Create a new .ivk request file. Specify the HTTP method, URL, and optionally headers, body, auth, scripts. The file is written to disk in .ivk format.
Inputs: file_path (required), method (required), url (required), name, +7 more.
Scan a folder for .ivk request files and .md documentation files. Returns the file tree with counts. Use this to discover what requests and docs exist in a collection.
Inputs: folder_path (required).
List all environments with their variables. Shows which environment is currently active. Use this to see available environments before switching or setting variables.
Inputs: no named parameters.
Read and parse an .ivk file. Returns the parsed request with method, URL, headers, body, directives, and scripts. Use this to inspect what a request does before sending it.
Inputs: file_path (required).
Execute an HTTP request from an .ivk file or inline params. Resolves {{variables}} from the active environment, runs pre/post scripts, and returns the full response with status, headers, body, test results, and console logs.
Inputs: file_path, method, url, headers, +1 more.
Set an environment variable in the active environment. The variable will be used to resolve {{name}} placeholders in subsequent requests.
Inputs: name (required), value (required).
Switch the active environment by name or index. All subsequent variable resolution will use the new environment's values.
Inputs: environment (required).
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 9 tools
Each tool has a clearly distinct purpose with no overlap: creating environments, building sites, creating/reading/sending requests, listing collections/environments, and managing variables. The descriptions clearly differentiate their functions, preventing misselection.
All tools follow a consistent 'invoker_verb_noun' pattern (e.g., invoker_add_environment, invoker_build_site). This predictable naming makes it easy for agents to understand and navigate the toolset without confusion.
With 9 tools, the server is well-scoped for managing Invoker collections, environments, and requests. Each tool serves a specific role in the workflow, from setup to execution, without being overly sparse or bloated.
The toolset covers core CRUD and lifecycle operations for environments, variables, and requests, with good workflow support. A minor gap exists in lacking tools for deleting or updating environments/requests, but agents can work around this with file system operations.
AI-callable tools for API mocking, testing, monitoring, security, and automation.
Discover, compare, route, and execute machine-accessible capabilities for AI agents.
Pay-per-call tools for AI agents, paid in USDC on Base or Solana. Web search, page and PDF reading, crypto news, Hyperliquid perps, market pulse, Chainlink price history, token checks, SEC filings, weather and more. Free calls every day, and a failed call costs nothing. Works with Tollkit Wallet.
Discover, inspect and run 63,000+ agent tools from one balance. Pay per call, no subscriptions.
