MCP NodeJS Debugger

Overview Schema Related Servers Score Discussions

nodejs_inspect

Execute JavaScript code directly in a debugged Node.js process to inspect variables, test expressions, and analyze runtime behavior during debugging sessions.

Instructions

Executes JavaScript code in the debugged process

Input Schema

TableJSON Schema

Name	Required	Description	Default
`js_code`	Yes	JavaScript code to execute

Implementation Reference

src/mcp-server.js:332-472 (registration)

Registration of the 'nodejs_inspect' tool using server.tool(), including description, input schema, and inline handler function.

server.tool(
  "nodejs_inspect",
  "Executes JavaScript code in the debugged process",
  {
    js_code: z.string().describe("JavaScript code to execute")
  },
  async ({ js_code }) => {
    try {
      // Ensure debugger is enabled
      if (!inspector.debuggerEnabled) {
        await inspector.enableDebugger();
      }
      
      // Capture the current console output length to know where to start capturing new output
      const consoleStartIndex = inspector.consoleOutput.length;
      
      // Wrap the code in a try-catch with explicit console logging for errors
      let codeToExecute = `
        try {
          ${js_code}
        } catch (e) {
          e;  // Return the error
        }
      `;
      
      const response = await inspector.send('Runtime.evaluate', {
        expression: codeToExecute,
        contextId: 1,
        objectGroup: 'console',
        includeCommandLineAPI: true,
        silent: false,
        returnByValue: true,
        generatePreview: true,
        awaitPromise: true  // This will wait for promises to resolve
      });
      
      // Give some time for console logs to be processed
      await new Promise(resolve => setTimeout(resolve, 200));
      
      // Get any console output that was generated during execution
      const consoleOutputs = inspector.consoleOutput.slice(consoleStartIndex);
      const consoleText = consoleOutputs.map(output => 
        `[${output.type}] ${output.message}`
      ).join('\n');
      
      // Process the return value
      let result;
      if (response.result) {
        if (response.result.type === 'object') {
          if (response.result.value) {
            // If we have a value, use it
            result = response.result.value;
          } else if (response.result.objectId) {
            // If we have an objectId but no value, the object was too complex to serialize directly
            // Get more details about the object
            try {
              const objectProps = await inspector.getProperties(response.result.objectId);
              const formattedObject = {};
              
              for (const prop of objectProps.result) {
                if (prop.value) {
                  if (prop.value.type === 'object' && prop.value.subtype !== 'null') {
                    // For nested objects, try to get their details too
                    if (prop.value.objectId) {
                      try {
                        const nestedProps = await inspector.getProperties(prop.value.objectId);
                        const nestedObj = {};
                        for (const nestedProp of nestedProps.result) {
                          if (nestedProp.value) {
                            if (nestedProp.value.value !== undefined) {
                              nestedObj[nestedProp.name] = nestedProp.value.value;
                            } else {
                              nestedObj[nestedProp.name] = nestedProp.value.description || 
                                `[${nestedProp.value.subtype || nestedProp.value.type}]`;
                            }
                          }
                        }
                        formattedObject[prop.name] = nestedObj;
                      } catch (nestedErr) {
                        formattedObject[prop.name] = prop.value.description || 
                          `[${prop.value.subtype || prop.value.type}]`;
                      }
                    } else {
                      formattedObject[prop.name] = prop.value.description || 
                        `[${prop.value.subtype || prop.value.type}]`;
                    }
                  } else if (prop.value.type === 'function') {
                    formattedObject[prop.name] = '[function]';
                  } else if (prop.value.value !== undefined) {
                    formattedObject[prop.name] = prop.value.value;
                  } else {
                    formattedObject[prop.name] = `[${prop.value.type}]`;
                  }
                }
              }
              
              result = formattedObject;
            } catch (propErr) {
              // If we can't get properties, at least show the object description
              result = response.result.description || `[${response.result.subtype || response.result.type}]`;
            }
          } else {
            // Fallback for objects without value or objectId
            result = response.result.description || `[${response.result.subtype || response.result.type}]`;
          }
        } else if (response.result.type === 'undefined') {
          result = undefined;
        } else if (response.result.value !== undefined) {
          result = response.result.value;
        } else {
          result = `[${response.result.type}]`;
        }
      }
      
      let responseContent = [];
      
      // Add console output if there was any
      if (consoleText.length > 0) {
        responseContent.push({
          type: "text", 
          text: `Console output:\n${consoleText}`
        });
      }
      
      // Add the result
      responseContent.push({
        type: "text",
        text: `Code executed successfully. Result: ${JSON.stringify(result, null, 2)}`
      });
      
      return { content: responseContent };
    } catch (err) {
      return {
        content: [{
          type: "text",
          text: `Error executing code: ${err.message}`
        }]
      };
    }
  }
);

src/mcp-server.js:338-471 (handler)

The handler function executes the provided JS code in the Node.js debugged process using the inspector, handles console output, serializes complex objects by fetching properties, and returns structured content.

async ({ js_code }) => {
  try {
    // Ensure debugger is enabled
    if (!inspector.debuggerEnabled) {
      await inspector.enableDebugger();
    }
    
    // Capture the current console output length to know where to start capturing new output
    const consoleStartIndex = inspector.consoleOutput.length;
    
    // Wrap the code in a try-catch with explicit console logging for errors
    let codeToExecute = `
      try {
        ${js_code}
      } catch (e) {
        e;  // Return the error
      }
    `;
    
    const response = await inspector.send('Runtime.evaluate', {
      expression: codeToExecute,
      contextId: 1,
      objectGroup: 'console',
      includeCommandLineAPI: true,
      silent: false,
      returnByValue: true,
      generatePreview: true,
      awaitPromise: true  // This will wait for promises to resolve
    });
    
    // Give some time for console logs to be processed
    await new Promise(resolve => setTimeout(resolve, 200));
    
    // Get any console output that was generated during execution
    const consoleOutputs = inspector.consoleOutput.slice(consoleStartIndex);
    const consoleText = consoleOutputs.map(output => 
      `[${output.type}] ${output.message}`
    ).join('\n');
    
    // Process the return value
    let result;
    if (response.result) {
      if (response.result.type === 'object') {
        if (response.result.value) {
          // If we have a value, use it
          result = response.result.value;
        } else if (response.result.objectId) {
          // If we have an objectId but no value, the object was too complex to serialize directly
          // Get more details about the object
          try {
            const objectProps = await inspector.getProperties(response.result.objectId);
            const formattedObject = {};
            
            for (const prop of objectProps.result) {
              if (prop.value) {
                if (prop.value.type === 'object' && prop.value.subtype !== 'null') {
                  // For nested objects, try to get their details too
                  if (prop.value.objectId) {
                    try {
                      const nestedProps = await inspector.getProperties(prop.value.objectId);
                      const nestedObj = {};
                      for (const nestedProp of nestedProps.result) {
                        if (nestedProp.value) {
                          if (nestedProp.value.value !== undefined) {
                            nestedObj[nestedProp.name] = nestedProp.value.value;
                          } else {
                            nestedObj[nestedProp.name] = nestedProp.value.description || 
                              `[${nestedProp.value.subtype || nestedProp.value.type}]`;
                          }
                        }
                      }
                      formattedObject[prop.name] = nestedObj;
                    } catch (nestedErr) {
                      formattedObject[prop.name] = prop.value.description || 
                        `[${prop.value.subtype || prop.value.type}]`;
                    }
                  } else {
                    formattedObject[prop.name] = prop.value.description || 
                      `[${prop.value.subtype || prop.value.type}]`;
                  }
                } else if (prop.value.type === 'function') {
                  formattedObject[prop.name] = '[function]';
                } else if (prop.value.value !== undefined) {
                  formattedObject[prop.name] = prop.value.value;
                } else {
                  formattedObject[prop.name] = `[${prop.value.type}]`;
                }
              }
            }
            
            result = formattedObject;
          } catch (propErr) {
            // If we can't get properties, at least show the object description
            result = response.result.description || `[${response.result.subtype || response.result.type}]`;
          }
        } else {
          // Fallback for objects without value or objectId
          result = response.result.description || `[${response.result.subtype || response.result.type}]`;
        }
      } else if (response.result.type === 'undefined') {
        result = undefined;
      } else if (response.result.value !== undefined) {
        result = response.result.value;
      } else {
        result = `[${response.result.type}]`;
      }
    }
    
    let responseContent = [];
    
    // Add console output if there was any
    if (consoleText.length > 0) {
      responseContent.push({
        type: "text", 
        text: `Console output:\n${consoleText}`
      });
    }
    
    // Add the result
    responseContent.push({
      type: "text",
      text: `Code executed successfully. Result: ${JSON.stringify(result, null, 2)}`
    });
    
    return { content: responseContent };
  } catch (err) {
    return {
      content: [{
        type: "text",
        text: `Error executing code: ${err.message}`
      }]
    };
  }
}

src/mcp-server.js:335-337 (schema)
Input schema for the tool, defining 'js_code' as a required string parameter.
```
{
  js_code: z.string().describe("JavaScript code to execute")
},
```

src/mcp-server.js:36-319 (helper)

The Inspector class is the core helper that manages connection to Node.js inspector protocol, sending CDP commands, handling events like paused/resumed/console, and providing utilities used by the nodejs_inspect handler.

class Inspector {
	constructor(port = 9229, retryOptions = { maxRetries: 5, retryInterval: 1000, continuousRetry: true }) {
		this.port = port;
		this.connected = false;
		this.pendingRequests = new Map();
		this.debuggerEnabled = false;
		this.breakpoints = new Map();
		this.paused = false;
		this.currentCallFrames = [];
		this.retryOptions = retryOptions;
		this.retryCount = 0;
		this.callbackHandlers = new Map();
		this.continuousRetryEnabled = retryOptions.continuousRetry;
		this.initialize();
	}

	async initialize() {
		try {
			// First, get the WebSocket URL from the inspector JSON API
			// Use 127.0.0.1 instead of localhost to avoid IPv6 issues
			const response = await fetch(`http://127.0.0.1:${this.port}/json`);
			const data = await response.json();
			const debuggerUrl = data[0]?.webSocketDebuggerUrl;
			
			if (!debuggerUrl) {
				this.scheduleRetry();
				return;
			}
			
			this.ws = new WebSocket(debuggerUrl);
			
			this.ws.on('open', () => {
				this.connected = true;
				this.retryCount = 0;
				this.enableDebugger();
			});
			
			this.ws.on('error', (error) => {
				this.scheduleRetry();
				 this.debuggerEnabled = false;
			});
			
			this.ws.on('close', () => {
				this.connected = false;
				this.scheduleRetry();
				this.debuggerEnabled = false;
			});
			
			this.ws.on('message', (data) => {
				const response = JSON.parse(data.toString());
				
				// Handle events
				if (response.method) {
					this.handleEvent(response);
					return;
				}
				
				// Handle response for pending request
				if (response.id && this.pendingRequests.has(response.id)) {
					const { resolve, reject } = this.pendingRequests.get(response.id);
					this.pendingRequests.delete(response.id);
					
					if (response.error) {
						reject(response.error);
					} else {
						resolve(response.result);
					}
				}
			});
		} catch (error) {
			this.scheduleRetry();
		}
	}
	
	scheduleRetry() {
		// If continuous retry is enabled, we'll keep trying after the initial attempts
		if (this.retryCount < this.retryOptions.maxRetries || this.continuousRetryEnabled) {
			this.retryCount++;
			
			// Use a longer interval for continuous retries to reduce resource usage
			const interval = this.continuousRetryEnabled && this.retryCount > this.retryOptions.maxRetries
				? Math.min(this.retryOptions.retryInterval * 5, 10000) // Max 10 seconds between retries
				: this.retryOptions.retryInterval;
				
			setTimeout(() => this.initialize(), interval);
		}
	}
	
	async enableDebugger() {
    try {
      if (!this.debuggerEnabled && this.connected) {

				await this.send('Debugger.enable', {});
				this.debuggerEnabled = true;
				
				// Setup event listeners
				await this.send('Runtime.enable', {});
				
				// Also activate possible domains we'll need
				await this.send('Runtime.runIfWaitingForDebugger', {});
			}
		} catch (error) {
			this.scheduleRetry();
    }
	}
	
	handleEvent(event) {
		
		switch (event.method) {
			case 'Debugger.paused':
				this.paused = true;
				this.currentCallFrames = event.params.callFrames;
				
				// Notify any registered callbacks for pause events
				if (this.callbackHandlers.has('paused')) {
					this.callbackHandlers.get('paused').forEach(callback => 
						callback(event.params));
				}
				break;
				
			case 'Debugger.resumed':
				this.paused = false;
				this.currentCallFrames = [];
				
				// Notify any registered callbacks for resume events
				if (this.callbackHandlers.has('resumed')) {
					this.callbackHandlers.get('resumed').forEach(callback => 
						callback());
				}
				break;
				
			case 'Debugger.scriptParsed':
				// Script parsing might be useful for source maps
				break;
				
			case 'Runtime.exceptionThrown':
				break;
				
			case 'Runtime.consoleAPICalled':
				// Handle console logs from the debugged program
				const args = event.params.args.map(arg => {
					if (arg.type === 'string') return arg.value;
					if (arg.type === 'number') return arg.value;
					if (arg.type === 'boolean') return arg.value;
					if (arg.type === 'object') {
						if (arg.value) {
							return JSON.stringify(arg.value, null, 2);
						} else if (arg.objectId) {
							// We'll try to get properties later as we can't do async here
							return arg.description || `[${arg.subtype || arg.type}]`;
						} else {
							return arg.description || `[${arg.subtype || arg.type}]`;
						}
					}
					return JSON.stringify(arg);
				}).join(' ');
				
				// Store console logs to make them available to the MCP tools
				if (!this.consoleOutput) {
					this.consoleOutput = [];
				}
				this.consoleOutput.push({
					type: event.params.type,
					message: args,
					timestamp: Date.now(),
					raw: event.params.args
				});
				
				// Keep only the last 100 console messages to avoid memory issues
				if (this.consoleOutput.length > 100) {
					this.consoleOutput.shift();
				}
				
				break;
		}
	}
	
	registerCallback(event, callback) {
		if (!this.callbackHandlers.has(event)) {
			this.callbackHandlers.set(event, []);
		}
		this.callbackHandlers.get(event).push(callback);
	}
	
	unregisterCallback(event, callback) {
		if (this.callbackHandlers.has(event)) {
			const callbacks = this.callbackHandlers.get(event);
			const index = callbacks.indexOf(callback);
			if (index !== -1) {
				callbacks.splice(index, 1);
			}
		}
	}

	async send(method, params) {
		return new Promise((resolve, reject) => {
			const timeout = setTimeout(() => {
				reject(new Error(`Request timed out: ${method}`));
				this.pendingRequests.delete(id);
			}, 5000);
			
			const checkConnection = () => {
				if (this.connected) {
					try {
						const id = Math.floor(Math.random() * 1000000);
						this.pendingRequests.set(id, { 
							resolve: (result) => {
								clearTimeout(timeout);
								resolve(result);
							}, 
							reject: (err) => {
								clearTimeout(timeout);
								reject(err);
							} 
						});
						
						this.ws.send(JSON.stringify({
							id,
							method,
							params
						}));
					} catch (err) {
						clearTimeout(timeout);
						reject(err);
					}
				} else {
					const connectionCheckTimer = setTimeout(checkConnection, 100);
					// If still not connected after 3 seconds, reject the promise
					setTimeout(() => {
						clearTimeout(connectionCheckTimer);
						clearTimeout(timeout);
						reject(new Error('Not connected to debugger'));
					}, 3000);
				}
			};
			
			checkConnection();
		});
	}
	
	async getScriptSource(scriptId) {
		try {
			const response = await this.send('Debugger.getScriptSource', {
				scriptId
			});
			return response.scriptSource;
		} catch (err) {
			return null;
		}
	}
	
	async evaluateOnCallFrame(callFrameId, expression) {
		if (!this.paused) {
			throw new Error('Debugger is not paused');
		}
		
		try {
			return await this.send('Debugger.evaluateOnCallFrame', {
				callFrameId,
				expression,
				objectGroup: 'console',
				includeCommandLineAPI: true,
				silent: false,
				returnByValue: true,
				generatePreview: true
			});
		} catch (err) {
			throw err;
		}
	}
	
	async getProperties(objectId, ownProperties = true) {
		try {
			return await this.send('Runtime.getProperties', {
				objectId,
				ownProperties,
				accessorPropertiesOnly: false,
				generatePreview: true
			});
		} catch (err) {
			throw err;
		}
	}
}

Tool Definition Quality

C2.9/5.0

Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description carries full burden but offers minimal behavioral insight. It mentions execution in a debugged process, hinting at a debugging context, but fails to disclose critical traits like safety implications (e.g., potential side effects on the process), authentication needs, or error handling. This leaves significant gaps for a tool that executes code.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single, efficient sentence with zero wasted words. It's front-loaded with the core action and context, making it easy to parse quickly without unnecessary elaboration.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the complexity of executing code in a debugged process, no annotations, and no output schema, the description is inadequate. It lacks details on return values, error cases, or behavioral nuances, leaving the agent with insufficient information to use the tool safely and effectively in a debugging context.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, with the single parameter 'js_code' well-documented in the schema. The description adds no additional meaning beyond implying execution context ('in the debugged process'), which doesn't enhance parameter understanding. This meets the baseline for high schema coverage.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the action ('Executes') and target ('JavaScript code in the debugged process'), providing a specific verb+resource combination. However, it doesn't explicitly distinguish this from sibling tools like 'evaluate' which might have similar debugging functions, preventing a perfect score.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides no guidance on when to use this tool versus alternatives like 'evaluate' or other debugging siblings. It lacks context about prerequisites (e.g., needing an active debug session) or exclusions, offering only a basic functional statement without usage direction.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Install Server

Other Tools

Latest Blog Posts

Tool Definition Quality Score (TDQS)
By punkpeye on April 3, 2026.
mcp
The Hackers Who Tracked My Sleep Cycle
By punkpeye on March 26, 2026.
security
Open Source Has a Bot Problem
By punkpeye on March 19, 2026.
open source

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/workbackai/mcp-nodejs-debugger'

If you have feedback or need assistance with the MCP directory API, please join our Discord server