Skip to main content
Glama

generate_report

Create structured Word documents from data payloads with Markdown formatting, tables, and sections for reports and documentation.

Instructions

Generates a complete Word document based on a structured content payload.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
filenameYesFilename for the generated report (e.g. 'audit_report')
titleYesTitle of the document
sectionsYesList of sections to include in the report

Implementation Reference

  • Core handler function that generates the Word report document by creating sections, paragraphs, tables using docx library, packing to buffer, and saving to output directory.
    public async generateReport(data: ReportData): Promise<string> {
        const doc = new Document({
            sections: [
                {
                    children: [
                        new Paragraph({
                            text: data.title,
                            heading: HeadingLevel.TITLE,
                            spacing: {
                                after: 200,
                            },
                        }),
                        ...this.renderSections(data.sections),
                    ],
                },
            ],
        });
    
        const buffer = await Packer.toBuffer(doc);
        const fileNameWithExt = data.filename.endsWith(".docx")
            ? data.filename
            : `${data.filename}.docx`;
        const filePath = path.join(this.outputDir, fileNameWithExt);
    
        fs.writeFileSync(filePath, buffer);
        return filePath;
    }
  • Zod schema defining the input structure for the generate_report tool, including filename, title, and sections with optional content and tables.
    const GenerateReportSchema = z.object({
        filename: z.string().describe("Filename for the generated report (e.g. 'audit_report')"),
        title: z.string().describe("Title of the document"),
        sections: z.array(
            z.object({
                heading: z.string(),
                content: z.string().optional().describe("Markdown-supported text content for this section"),
                table: z.object({
                    headers: z.array(z.string()),
                    rows: z.array(z.array(z.string())),
                }).optional(),
            })
        ).describe("List of sections to include in the report"),
    }).describe("Generates a complete Word document based on a structured content payload.");
  • src/index.ts:39-68 (registration)
    Registers the generate_report tool with the MCP server, providing name, description, input schema, and a handler that delegates to WordGenerator.generateReport and returns success/error message.
    server.tool(
        "generate_report",
        "Generates a complete Word document based on a structured content payload.",
        GenerateReportSchema.shape,
        async (args) => {
            try {
                console.error(`Generating report: ${args.filename}`); // Log to stderr for debugging
                const filePath = await wordGenerator.generateReport(args);
                return {
                    content: [
                        {
                            type: "text",
                            text: `Report generated successfully at ${filePath}`
                        }
                    ],
                };
            } catch (error) {
                console.error("Error generating report:", error);
                return {
                    content: [
                        {
                            type: "text",
                            text: `Error generating report: ${error instanceof Error ? error.message : String(error)}`
                        }
                    ],
                    isError: true,
                };
            }
        }
    );
  • TypeScript interface defining the ReportData type used by the generateReport handler, matching the tool schema.
    export interface ReportData {
        filename: string;
        title: string;
        sections: ReportSection[];
    }
  • Helper method to render report sections into docx Paragraph and Table elements, handling headings, markdown content, and tables.
    private renderSections(sections: ReportSection[]): (Paragraph | Table)[] {
        const elements: (Paragraph | Table)[] = [];
    
        for (const section of sections) {
            // detailed heading
            elements.push(
                new Paragraph({
                    text: section.heading,
                    heading: HeadingLevel.HEADING_1,
                    spacing: {
                        before: 200,
                        after: 100,
                    },
                })
            );
    
            // plain content (basic markdown support: splitting by lines)
            if (section.content) {
                const lines = section.content.split("\n");
                for (const line of lines) {
                    // Basic bold handling for **text**
                    // This is a very primitive parser.
                    const textRuns = this.parseMarkdownLine(line);
                    elements.push(
                        new Paragraph({
                            children: textRuns,
                            spacing: {
                                after: 100
                            }
                        })
                    );
                }
            }
    
            // table
            if (section.table) {
                elements.push(this.createTable(section.table.headers, section.table.rows));
            }
        }
    
        return elements;
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/MokhtarLahjaily/word-mcp'

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