Docker MCP Server

import { describe, it, expect, beforeAll, afterAll, beforeEach } from '@jest/globals'; import { Client } from '@modelcontextprotocol/sdk/client/index.js'; import { createTestClient, randomFilename, writeFile, readFile, cleanupTestFiles } from '../helpers.js'; describe('Line Truncation', () => { let client: Client; beforeAll(async () => { client = await createTestClient(); }); afterAll(async () => { if (client) { await cleanupTestFiles(client); await client.close(); } }); beforeEach(async () => { await cleanupTestFiles(client); }); describe('file_read truncation', () => { it('should truncate lines longer than 2000 characters', async () => { const filename = randomFilename(); // Create a line with 3000 'a' characters const longLine = 'a'.repeat(3000); const content = `short line\n${longLine}\nanother short line`; await writeFile(client, filename, content); const result = await readFile(client, filename); // The long line should be truncated to 2000 chars // Since cat -n adds line numbers, we check that we don't see 3000 a's const lines = result.split('\n'); const longLineResult = lines.find(l => l.includes('aaa')); expect(longLineResult).toBeDefined(); // Count the 'a' characters - should be at most 2000 const aCount = (longLineResult!.match(/a/g) || []).length; expect(aCount).toBeLessThanOrEqual(2000); expect(aCount).toBeGreaterThan(1000); // Should still have substantial content }); it('should not truncate lines under 2000 characters', async () => { const filename = randomFilename(); const normalLine = 'x'.repeat(500); const content = normalLine; await writeFile(client, filename, content); const result = await readFile(client, filename); // All 500 x's should be present const xCount = (result.match(/x/g) || []).length; expect(xCount).toBe(500); }); }); describe('file_grep truncation', () => { it('should truncate matched lines longer than 2000 characters', async () => { const filename = randomFilename(); // Create a line with 3000 characters that contains a searchable pattern const longLine = 'FINDME' + 'b'.repeat(2994); const content = `short line\n${longLine}\nanother line`; await writeFile(client, filename, content); const result = await client.callTool({ name: 'file_grep', arguments: { pattern: 'FINDME', path: '.' } }); const text = (result.content as Array<{text: string}>)[0].text; // The matched line should be truncated // Count b's - should be less than 2994 const bCount = (text.match(/b/g) || []).length; expect(bCount).toBeLessThan(2994); expect(text).toContain('FINDME'); }); it('should not truncate grep results under 200 characters', async () => { const filename = randomFilename(); // grep truncates at 200 chars, so use a line that fits within that limit // SEARCHTERM (10 chars) + 100 z's = 110 chars total, well under 200 // Using 'z' to avoid matching random chars in filename const normalLine = 'SEARCHTERM' + 'z'.repeat(100); const content = normalLine; await writeFile(client, filename, content); const result = await client.callTool({ name: 'file_grep', arguments: { pattern: 'SEARCHTERM', path: '.' } }); const text = (result.content as Array<{text: string}>)[0].text; // All 100 z's should be present (line is under 200 char limit) const zCount = (text.match(/z/g) || []).length; expect(zCount).toBe(100); }); }); });