Merge pull request #4 from runloopai/rl-cli-python-replace

Command line capabilities

Author: dines-rl

.gitignore

@@ -3,3 +3,4 @@ dist/
 *.log
 .DS_Store
 .env
+coverage

CLAUDE_SETUP.md

@@ -5,15 +5,15 @@ This guide will walk you through connecting the Runloop MCP server to Claude Des
 ## Prerequisites
 
 1. Make sure you have Claude Desktop installed
-2. Authenticate with Runloop: `rln auth`
-3. Make sure `rln` is installed globally and in your PATH
+2. Authenticate with Runloop: `rli auth`
+3. Make sure `rli` is installed globally and in your PATH
 
 ## Quick Setup (Automatic)
 
 The easiest way to set up Runloop with Claude Desktop:
 
 ```bash
-rln mcp install
+rli mcp install
 ```
 
 This command will:
@@ -54,7 +54,7 @@ If the file doesn't exist, create it. Add or update it with this configuration:
 {
   "mcpServers": {
     "runloop": {
-      "command": "rln",
+      "command": "rli",
       "args": ["mcp", "start"]
     }
   }
@@ -71,7 +71,7 @@ If the file doesn't exist, create it. Add or update it with this configuration:
       "args": ["some-args"]
     },
     "runloop": {
-      "command": "rln",
+      "command": "rli",
       "args": ["mcp", "start"]
     }
   }
@@ -98,17 +98,17 @@ Claude will now have access to these Runloop tools and can manage your devboxes!
 
 ## Troubleshooting
 
-### "Command not found: rln"
+### "Command not found: rli"
 
-Make sure `rln` is in your PATH. Test by running `which rln` (macOS/Linux) or `where rln` (Windows) in your terminal.
+Make sure `rli` is in your PATH. Test by running `which rli` (macOS/Linux) or `where rli` (Windows) in your terminal.
 
 If not found:
 - If installed via npm globally: `npm install -g @runloop/rl-cli`
 - Check your npm global bin directory is in PATH: `npm config get prefix`
 
 ### "API key not configured"
 
-Run `rln auth` to configure your API key before using the MCP server.
+Run `rli auth` to configure your API key before using the MCP server.
 
 ### Claude doesn't show Runloop tools
 
@@ -123,7 +123,7 @@ Run `rln auth` to configure your API key before using the MCP server.
 You can test if the MCP server is working by running:
 
 ```bash
-rln mcp start
+rli mcp start
 ```
 
 It should output: `Runloop MCP server running on stdio`
@@ -140,7 +140,7 @@ If you want to connect to Runloop's development environment:
 {
   "mcpServers": {
     "runloop": {
-      "command": "rln",
+      "command": "rli",
       "args": ["mcp", "start"],
       "env": {
         "RUNLOOP_ENV": "dev"
@@ -152,20 +152,20 @@ If you want to connect to Runloop's development environment:
 
 ### Using a Specific Path
 
-If `rln` isn't in your PATH, you can specify the full path:
+If `rli` isn't in your PATH, you can specify the full path:
 
 ```json
 {
   "mcpServers": {
     "runloop": {
-      "command": "/full/path/to/rln",
+      "command": "/full/path/to/rli",
       "args": ["mcp", "start"]
     }
   }
 }
 ```
 
-Find the full path with: `which rln` (macOS/Linux) or `where rln` (Windows)
+Find the full path with: `which rli` (macOS/Linux) or `where rli` (Windows)
 
 ## What Can Claude Do Now?
 

MCP_COMMANDS.md

@@ -9,7 +9,7 @@ Quick reference for all Runloop MCP commands.
 Automatically configure Claude Desktop to use Runloop MCP:
 
 ```bash
-rln mcp install
+rli mcp install
 ```
 
 This command:
@@ -27,7 +27,7 @@ After running this, restart Claude Desktop.
 Start the MCP server in stdio mode:
 
 ```bash
-rln mcp start
+rli mcp start
 ```
 
 This mode is used by Claude Desktop and other local AI assistants. The server communicates via stdin/stdout.
@@ -37,7 +37,7 @@ This mode is used by Claude Desktop and other local AI assistants. The server co
 Start the MCP server in HTTP mode:
 
 ```bash
-rln mcp start --http
+rli mcp start --http
 ```
 
 This starts an HTTP server on port 3000 by default, using Server-Sent Events (SSE) for communication.
@@ -47,18 +47,18 @@ This starts an HTTP server on port 3000 by default, using Server-Sent Events (SS
 Start the HTTP server on a specific port:
 
 ```bash
-rln mcp start --http --port 8080
+rli mcp start --http --port 8080
 ```
 
 ## Configuration File Format
 
-When you run `rln mcp install`, it creates this configuration in your Claude Desktop config:
+When you run `rli mcp install`, it creates this configuration in your Claude Desktop config:
 
 ```json
 {
   "mcpServers": {
     "runloop": {
-      "command": "rln",
+      "command": "rli",
       "args": ["mcp", "start"]
     }
   }
@@ -73,7 +73,7 @@ To use the development environment:
 {
   "mcpServers": {
     "runloop": {
-      "command": "rln",
+      "command": "rli",
       "args": ["mcp", "start"],
       "env": {
         "RUNLOOP_ENV": "dev"
@@ -120,15 +120,15 @@ Once set up, you can ask Claude:
 Test the stdio server manually:
 
 ```bash
-rln mcp start
+rli mcp start
 ```
 
 You should see: `Runloop MCP server running on stdio`
 
 Test the HTTP server:
 
 ```bash
-rln mcp start --http
+rli mcp start --http
 ```
 
 You should see:
@@ -142,19 +142,19 @@ Message endpoint: http://localhost:3000/message
 
 ### Command not found
 
-If you get "command not found: rln":
+If you get "command not found: rli":
 - Install globally: `npm install -g @runloop/rl-cli`
 - Check your PATH includes npm global bin directory
 
 ### API key not configured
 
-Run `rln auth` before using the MCP server.
+Run `rli auth` before using the MCP server.
 
 ### Port already in use
 
 For HTTP mode, use a different port:
 ```bash
-rln mcp start --http --port 8080
+rli mcp start --http --port 8080
 ```
 
 ## See Also

MCP_README.md

@@ -15,7 +15,7 @@ Runloop provides two transport modes for the MCP server:
 For local AI assistants like Claude Desktop:
 
 ```bash
-rln mcp start
+rli mcp start
 ```
 
 The server runs on stdio and communicates using the MCP protocol.
@@ -25,13 +25,13 @@ The server runs on stdio and communicates using the MCP protocol.
 For web-based AI assistants or remote access:
 
 ```bash
-rln mcp start --http
+rli mcp start --http
 ```
 
 Or specify a custom port:
 
 ```bash
-rln mcp start --http --port 8080
+rli mcp start --http --port 8080
 ```
 
 The HTTP server runs on `http://localhost:3000` by default and uses Server-Sent Events (SSE) for communication.
@@ -83,7 +83,7 @@ The MCP server exposes the following tools:
 The easiest way to set up with Claude Desktop:
 
 ```bash
-rln mcp install
+rli mcp install
 ```
 
 This automatically adds the configuration to your Claude Desktop config file and preserves any existing MCP servers.
@@ -100,7 +100,7 @@ If you prefer to configure manually, add this to your Claude configuration file:
 {
   "mcpServers": {
     "runloop": {
-      "command": "rln",
+      "command": "rli",
       "args": ["mcp", "start"],
       "env": {
         "RUNLOOP_ENV": "prod"
@@ -134,7 +134,7 @@ Example for Claude Code or other MCP clients supporting HTTP:
 The MCP server uses the same API key configuration as the CLI. Make sure you've authenticated first:
 
 ```bash
-rln auth
+rli auth
 ```
 
 The server will automatically use your stored API credentials.
@@ -154,32 +154,32 @@ Claude will use the MCP tools to interact with your Runloop account and provide
 ## Environment Variables
 
 - `RUNLOOP_ENV` - Set to `dev` for development environment, `prod` (or leave unset) for production
-- API key is read from the CLI configuration (~/.config/rln/config.json)
+- API key is read from the CLI configuration (~/.config/rli/config.json)
 
 ## Troubleshooting
 
 ### Stdio Server
 
 If the stdio MCP server isn't working:
 
-1. Make sure you've run `rln auth` to configure your API key
-2. Check that the `rln` command is in your PATH
+1. Make sure you've run `rli auth` to configure your API key
+2. Check that the `rli` command is in your PATH
 3. Restart Claude Desktop after updating the configuration
 4. Check Claude's logs for any error messages
 
 ### HTTP Server
 
 If the HTTP MCP server isn't working:
 
-1. Make sure you've run `rln auth` to configure your API key
+1. Make sure you've run `rli auth` to configure your API key
 2. Check that the port isn't already in use
 3. Verify the server is running: `curl http://localhost:3000/sse`
 4. Check your firewall settings if connecting remotely
 5. Look at the server logs for error messages
 
 ### Common Issues
 
-- **"API key not configured"**: Run `rln auth` to set up your credentials
+- **"API key not configured"**: Run `rli auth` to set up your credentials
 - **Port already in use**: Stop other services or use a different port with `--port`
 - **Connection refused**: Make sure the server is running and accessible