Skip to main content
Glama

execute_background_script

Execute server-side JavaScript in ServiceNow using Background Scripts to run custom code, automate processes, and interact with instance data within specified application scopes.

Instructions

Execute server-side JavaScript in ServiceNow using Background Scripts. ⚠️ SANDBOX ONLY - executes arbitrary code. 🛡️ Auto-truncates large outputs. 📁 Use {{file:path}} for large scripts.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
scriptYesThe server-side JavaScript code to execute (e.g., gs.print("Hello");). Maximum 50,000 characters. Supports {{file:...}} placeholders to load content from local files.
scopeYesThe application scope (e.g., "global" or specific app scope). Required.
timeout_msNoOptional timeout in milliseconds (default: 60000, range: 1000-300000)
include_htmlNoInclude HTML output in response (default: true). Set to false for text-only mode to reduce response size.
response_modeNoResponse verbosity: full (all data), minimal (essential only), compact (summarized). Default: full

Implementation Reference

  • MCP CallToolRequest handler specifically for 'execute_background_script': initializes client check, extracts and validates parameters (script, scope, etc.), resolves file placeholders, executes via client.executeScript, applies overflow prevention, formats JSON response, handles all errors.
    if (request.params.name === 'execute_background_script') {
      try {
        // Check if client was initialized successfully
        if (!client) {
          return {
            content: [
              {
                type: 'text',
                text: JSON.stringify({
                  success: false,
                  error: {
                    code: 'INITIALIZATION_ERROR',
                    message: 'ServiceNow client failed to initialize',
                    details:
                      initError instanceof Error ? initError.message : 'Unknown error',
                  },
                } as ServiceNowErrorResponse),
              },
            ],
            isError: true,
          };
        }
    
        // Extract parameters from tool arguments
        const args = request.params.arguments as Record<string, unknown> | undefined;
        const script = args?.script as string | undefined;
        const scope = args?.scope as string | undefined;
        const timeoutMs = args?.timeout_ms as number | undefined;
        const includeHtml = args?.include_html as boolean | undefined;
        const responseMode = args?.response_mode as string | undefined;
    
        // Validate required parameters
        if (!script) {
          return {
            content: [
              {
                type: 'text',
                text: JSON.stringify({
                  success: false,
                  error: {
                    code: ERROR_CODES.MISSING_PARAMETER,
                    message: 'Required parameter "script" is missing',
                    details: 'Please provide the JavaScript code to execute',
                  },
                } as ServiceNowErrorResponse),
              },
            ],
            isError: true,
          };
        }
    
        if (!scope) {
          return {
            content: [
              {
                type: 'text',
                text: JSON.stringify({
                  success: false,
                  error: {
                    code: ERROR_CODES.MISSING_PARAMETER,
                    message: 'Required parameter "scope" is missing',
                    details: 'Please provide the application scope (e.g., "global")',
                  },
                } as ServiceNowErrorResponse),
              },
            ],
            isError: true,
          };
        }
    
        // Resolve file placeholders in the request
        let resolvedScript = script;
        let resolvedScope = scope;
        
        try {
          const scriptResolution = resolveFilePlaceholders(script);
          resolvedScript = scriptResolution.data;
          
          if (scope) {
            const scopeResolution = resolveFilePlaceholders(scope);
            resolvedScope = scopeResolution.data;
          }
        } catch (error) {
          if (error instanceof FilePlaceholderError) {
            return {
              content: [
                {
                  type: 'text',
                  text: JSON.stringify({
                    success: false,
                    error: {
                      code: 'FILE_PLACEHOLDER_ERROR',
                      message: 'Failed to resolve file placeholder',
                      details: `${error.placeholder}: ${error.message}`,
                    },
                  } as ServiceNowErrorResponse),
                },
              ],
              isError: true,
            };
          }
          
          // Re-throw unknown errors
          throw error;
        }
    
        // Execute the script
        const result = await client.executeScript({
          script: resolvedScript,
          scope: resolvedScope,
          timeoutMs,
          include_html: includeHtml,
          response_mode: responseMode as any,
        });
    
        // Apply global context overflow prevention
        const { response: protectedResult, monitoring } = globalContextOverflowPrevention.monitorResponse(result, 'execute_background_script', responseMode);
    
        return {
          content: [
            {
              type: 'text',
              text: JSON.stringify(protectedResult, null, 2),
            },
          ],
        };
    
      } catch (error) {
        const errorResponse: ServiceNowErrorResponse = {
          success: false,
          error: {
            code: 'UNKNOWN_ERROR',
            message: 'An unexpected error occurred',
          },
        };
    
        if (error instanceof ServiceNowScriptError) {
          errorResponse.error.code = error.code;
          errorResponse.error.message = error.message;
          errorResponse.error.details = `HTTP Status: ${error.statusCode || 'N/A'}`;
        } else if (error instanceof Error) {
          errorResponse.error.message = error.message;
          errorResponse.error.details = 'Check ServiceNow instance connectivity and credentials';
        }
    
        return {
          content: [
            {
              type: 'text',
              text: JSON.stringify(errorResponse),
            },
          ],
          isError: true,
        };
      }
    }
  • Tool schema definition in ListToolsRequest handler: specifies name, detailed description, and inputSchema with properties, types, descriptions, and required fields for the execute_background_script tool.
      name: 'execute_background_script',
      description:
        'Execute server-side JavaScript in ServiceNow using Background Scripts. ⚠️ SANDBOX ONLY - executes arbitrary code. 🛡️ Auto-truncates large outputs. 📁 Use {{file:path}} for large scripts.',
      inputSchema: {
        type: 'object',
        properties: {
          script: {
            type: 'string',
            description:
              'The server-side JavaScript code to execute (e.g., gs.print("Hello");). Maximum 50,000 characters. Supports {{file:...}} placeholders to load content from local files.',
          },
          scope: {
            type: 'string',
            description:
              'The application scope (e.g., "global" or specific app scope). Required.',
          },
          timeout_ms: {
            type: 'number',
            description:
              'Optional timeout in milliseconds (default: 60000, range: 1000-300000)',
          },
          include_html: {
            type: 'boolean',
            description:
              'Include HTML output in response (default: true). Set to false for text-only mode to reduce response size.',
          },
          response_mode: {
            type: 'string',
            enum: ['full', 'minimal', 'compact'],
            description:
              'Response verbosity: full (all data), minimal (essential only), compact (summarized). Default: full',
          },
        },
        required: ['script', 'scope'],
      },
    },
  • src/index.ts:583-790 (registration)
    Registration of all MCP tools via setRequestHandler(ListToolsRequestSchema), listing 'execute_background_script' with its schema and description.
    server.setRequestHandler(ListToolsRequestSchema, async () => {
      return {
        tools: [
          {
            name: 'execute_background_script',
            description:
              'Execute server-side JavaScript in ServiceNow using Background Scripts. ⚠️ SANDBOX ONLY - executes arbitrary code. 🛡️ Auto-truncates large outputs. 📁 Use {{file:path}} for large scripts.',
            inputSchema: {
              type: 'object',
              properties: {
                script: {
                  type: 'string',
                  description:
                    'The server-side JavaScript code to execute (e.g., gs.print("Hello");). Maximum 50,000 characters. Supports {{file:...}} placeholders to load content from local files.',
                },
                scope: {
                  type: 'string',
                  description:
                    'The application scope (e.g., "global" or specific app scope). Required.',
                },
                timeout_ms: {
                  type: 'number',
                  description:
                    'Optional timeout in milliseconds (default: 60000, range: 1000-300000)',
                },
                include_html: {
                  type: 'boolean',
                  description:
                    'Include HTML output in response (default: true). Set to false for text-only mode to reduce response size.',
                },
                response_mode: {
                  type: 'string',
                  enum: ['full', 'minimal', 'compact'],
                  description:
                    'Response verbosity: full (all data), minimal (essential only), compact (summarized). Default: full',
                },
              },
              required: ['script', 'scope'],
            },
          },
          {
            name: 'execute_table_operation',
            description:
              'CRUD operations on ServiceNow tables via Table API. Supports GET/POST/PUT/PATCH/DELETE with query syntax and batch operations. ⚠️ SANDBOX ONLY - reads/modifies data. 🛡️ Auto-limits large results. Use pagination for big datasets. 📁 Use {{file:path}} for large data.',
            inputSchema: {
              type: 'object',
              properties: {
                operation: {
                  type: 'string',
                  enum: ['GET', 'POST', 'PUT', 'PATCH', 'DELETE'],
                  description: 'The operation to perform on the table. Required.',
                },
                table: {
                  type: 'string',
                  description: 'The ServiceNow table name (e.g., "incident", "sys_user"). Required.',
                },
                sys_id: {
                  type: 'string',
                  description: 'System ID for single record operations (GET, PUT, PATCH, DELETE).',
                },
                sys_ids: {
                  type: 'array',
                  items: { type: 'string' },
                  description: 'Array of system IDs for batch operations.',
                },
                query: {
                  type: 'string',
                  description: 'ServiceNow encoded query string (e.g., "active=true^priority=1").',
                },
                fields: {
                  type: 'string',
                  description: 'Comma-separated list of fields to return.',
                },
                limit: {
                  type: 'number',
                  description: 'Maximum number of records to return (default: 1000).',
                },
                offset: {
                  type: 'number',
                  description: 'Number of records to skip for pagination.',
                },
                display_value: {
                  type: 'string',
                  enum: ['true', 'false', 'all'],
                  description: 'Return display values for reference fields.',
                },
                exclude_reference_link: {
                  type: 'boolean',
                  description: 'Exclude reference link fields from response.',
                },
                data: {
                  type: 'object',
                  description: 'Record data for POST/PUT/PATCH operations. Can be single object or array for batch operations. Supports {{file:...}} placeholders to load content from local files.',
                },
                batch: {
                  type: 'boolean',
                  description: 'Enable batch mode for multiple record operations.',
                },
                validate_fields: {
                  type: 'boolean',
                  description: 'Enable field validation warnings to catch typos and invalid field names. Default: true (validation enabled by default).',
                },
                context_overflow_prevention: {
                  type: 'boolean',
                  description: 'Enable context overflow prevention to limit large result sets. Default: true. Set to false to disable automatic truncation (use with caution).',
                },
                strict_fields: {
                  type: 'boolean',
                  description: 'Strict field filtering - only return requested fields and strip large fields (script, html, css) unless explicitly requested. Default: false.',
                },
                response_mode: {
                  type: 'string',
                  enum: ['full', 'minimal', 'compact'],
                  description: 'Response verbosity: full (all data), minimal (essential only), compact (summarized). Default: full',
                },
              },
              required: ['operation', 'table'],
            },
          },
          {
            name: 'execute_updateset_operation',
            description:
              'Manage ServiceNow update sets with lifecycle operations, XML reassignment, and working set tracking. ⚠️ SANDBOX ONLY - modifies update sets. 🛡️ Auto-limits large results. Use pagination for big datasets. 📁 Use {{file:path}} for large data.',
            inputSchema: {
              type: 'object',
              properties: {
                operation: {
                  type: 'string',
                  enum: ['create', 'set_working', 'show_working', 'clear_working', 'insert', 'update', 'rehome', 'contents', 'recent', 'list', 'info', 'complete', 'reopen', 'delete', 'diff_default'],
                  description: 'The update set operation to perform. Required.',
                },
                name: {
                  type: 'string',
                  description: 'Update set name (required for create operation).',
                },
                description: {
                  type: 'string',
                  description: 'Update set description (optional for create operation).',
                },
                scope: {
                  type: 'string',
                  description: 'Update set scope (optional, defaults to configured scope).',
                },
                set_as_working: {
                  type: 'boolean',
                  description: 'Set the created update set as working set (for create operation).',
                },
                update_set_sys_id: {
                  type: 'string',
                  description: 'Update set sys_id for operations that require it.',
                },
                table: {
                  type: 'string',
                  description: 'Table name for insert/update operations.',
                },
                sys_id: {
                  type: 'string',
                  description: 'Record sys_id for update operations.',
                },
                data: {
                  type: 'object',
                  description: 'Record data for insert/update operations. Can be single object or array for batch operations. Supports {{file:...}} placeholders to load content from local files.',
                },
                batch: {
                  type: 'boolean',
                  description: 'Enable batch mode for multiple record operations.',
                },
                xml_sys_ids: {
                  type: 'array',
                  items: { type: 'string' },
                  description: 'Array of XML sys_ids for rehome operations.',
                },
                query: {
                  type: 'string',
                  description: 'ServiceNow encoded query string for rehome operations.',
                },
                force: {
                  type: 'boolean',
                  description: 'Force reassignment even if XML is not in Default update set.',
                },
                limit: {
                  type: 'number',
                  description: 'Maximum number of records to return for list/recent operations.',
                },
                offset: {
                  type: 'number',
                  description: 'Number of records to skip for pagination.',
                },
                filters: {
                  type: 'object',
                  description: 'Filters for list operations (scope, state, created_by, sys_created_on).',
                },
                response_mode: {
                  type: 'string',
                  enum: ['full', 'minimal', 'compact'],
                  description: 'Response verbosity: full (all data), minimal (essential only), compact (summarized). Default: full',
                },
                quiet: {
                  type: 'boolean',
                  description: 'Compact acknowledgment for update operations to avoid RESPONSE_TOO_LARGE errors. Default: false.',
                },
              },
              required: ['operation'],
            },
          },
        ],
      };
    });
  • Core helper method ServiceNowBackgroundScriptClient.executeScript(): performs full flow of authenticating UI session, fetching/parsing background script form, posting script execution with retry on session timeout, parsing HTML/text output, returning structured ScriptExecutionResult.
    async executeScript(request: ScriptExecutionRequest): Promise<ScriptExecutionResult> {
      const startTime = Date.now();
    
      try {
    
        // Validate input
        this.validateScriptRequest(request);
    
        // Step 1: Establish UI session via form login
        await this.establishUISession();
    
        // Step 2: GET the background script page and parse form data
        const pageResponse = await this.getBackgroundScriptPage();
        const { actionUrl, formData } = this.parseFormData(pageResponse.html, request.script, request.scope);
    
        // Step 3: POST the script execution request with retry logic
        let executionResponse = await this.postScriptExecution({
          actionUrl,
          formData,
          timeoutMs: request.timeoutMs || this.clientConfig.timeoutMs,
        });
    
        // Log telemetry (expert specification)
        const cookieNames = this.getCookieHeader(actionUrl).split(';').map(c => c.split('=')[0]).filter(c => c).join(',');
    
        // Check for session timeout and retry if needed
        let retry = false;
        if (this.hasSessionTimeout(executionResponse.body)) {
          retry = true;
          
          // Get fresh page and parse form data
          const freshPageResponse = await this.getBackgroundScriptPage();
          const { actionUrl: freshActionUrl, formData: freshFormData } = this.parseFormData(freshPageResponse.html, request.script, request.scope);
          
          // Retry POST with fresh form data
          executionResponse = await this.postScriptExecution({
            actionUrl: freshActionUrl,
            formData: freshFormData,
            timeoutMs: request.timeoutMs || this.clientConfig.timeoutMs,
          });
          
        }
    
        // Step 3: Parse the response
        const parsedOutput = parseScriptOutput(executionResponse.body);
        const executionTime = Date.now() - startTime;
        
        // Log pre blocks telemetry
        const preBlocks = (executionResponse.body.match(/<pre[^>]*>/gi) || []).length;
    
        // Handle response_mode shortcuts and include_html parameter for text-only mode
        const shouldIncludeHtml = request.include_html !== false && request.response_mode !== 'minimal';
        
        if (!shouldIncludeHtml) {
          return {
            success: true,
            output: {
              text: parsedOutput.text,
            },
            metadata: {
              executionTime,
              htmlSize: 0,
              timestamp: new Date().toISOString(),
              scope: request.scope,
            },
          };
        }
    
        return {
          success: true,
          output: {
            html: parsedOutput.html,
            text: parsedOutput.text,
          },
          metadata: {
            executionTime,
            htmlSize: parsedOutput.html.length,
            timestamp: new Date().toISOString(),
            scope: request.scope,
          },
        };
    
      } catch (error) {
        if (error instanceof ServiceNowScriptError) {
          throw error;
        }
        
        throw new ServiceNowScriptError(
          ERROR_CODES.SCRIPT_EXECUTION_ERROR,
          undefined,
          `Script execution failed: ${error instanceof Error ? error.message : 'Unknown error'}`
        );
      }
    }
Behavior5/5

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

With no annotations provided, the description carries the full burden of behavioral disclosure and does so exceptionally well. It explicitly warns about the sandbox environment ('⚠️ SANDBOX ONLY - executes arbitrary code'), discloses output handling behavior ('🛡️ Auto-truncates large outputs'), and provides practical implementation guidance. This goes well beyond what the input schema provides about parameters.

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 perfectly sized and front-loaded with the core purpose. Every sentence earns its place: the first states the primary function, the second provides critical warnings, and the third offers practical implementation advice. The emoji usage enhances scannability without adding fluff.

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

Completeness4/5

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

For a complex tool that executes arbitrary code with 5 parameters and no output schema, the description provides excellent context about behavior, warnings, and practical usage. The only minor gap is the lack of information about return values or error handling, which would be helpful given there's no output schema. However, the description covers the most critical aspects well.

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?

With 100% schema description coverage, the input schema already thoroughly documents all 5 parameters. The description adds some value by reinforcing the script parameter's support for file placeholders, but doesn't provide additional semantic context beyond what's already in the schema descriptions. This meets the baseline expectation when schema coverage is complete.

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

Purpose5/5

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

The description clearly states the specific action ('Execute server-side JavaScript in ServiceNow using Background Scripts') and resource ('ServiceNow'), distinguishing it from sibling tools like execute_table_operation and execute_updateset_operation which likely operate on different resources. It provides a precise verb+resource combination that leaves no ambiguity about what the tool does.

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

Usage Guidelines4/5

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

The description provides clear context for when to use this tool ('Execute server-side JavaScript in ServiceNow') and includes practical guidance for handling large scripts ('Use {{file:path}} for large scripts'). However, it doesn't explicitly state when NOT to use it or directly compare it to sibling alternatives, which prevents a perfect score.

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

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/ClearSkye/SkyeNet-MCP-ACE'

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