mcp-adr-analysis-server

Instructions

Implementation Reference

• src/tools/content-masking-tool.ts:870-1069 (handler)

  The primary handler function for the 'configure_custom_patterns' tool. It performs actual filesystem scans of the project to gather package.json, .env, config, and script files, then generates detailed prompts for AI to create project-specific custom security masking patterns. Returns structured MCP response with instructions and prompts.
  export async function configureCustomPatterns(args: { projectPath: string; existingPatterns?: string[]; }): Promise<any> { const { projectPath, existingPatterns } = args; try { // Use actual file operations to scan project structure const { scanProjectStructure } = await import('../utils/actual-file-operations.js'); // Actually scan project structure const projectStructure = await scanProjectStructure(projectPath, { readContent: true, maxFileSize: 10000, }); const customPatternPrompt = ` # Custom Pattern Configuration Generation Based on actual project structure analysis, here are the findings: ## Project Structure - **Root Path**: ${projectStructure.rootPath} - **Total Files**: ${projectStructure.totalFiles} - **Directories**: ${projectStructure.directories.join(', ')} ## Package Management Files ${ projectStructure.packageFiles.length > 0 ? projectStructure.packageFiles .map( f => ` ### ${f.filename} \`\`\` ${f.content.slice(0, 500)}${f.content.length > 500 ? '\n... (truncated)' : ''} \`\`\` ` ) .join('\n') : '- No package files found' } ## Environment Configuration Files ${ projectStructure.environmentFiles.length > 0 ? projectStructure.environmentFiles .map( f => ` ### ${f.filename} \`\`\` ${f.content.slice(0, 300)}${f.content.length > 300 ? '\n... (truncated)' : ''} \`\`\` ` ) .join('\n') : '- No environment files found' } ## Configuration Files ${ projectStructure.configFiles.length > 0 ? projectStructure.configFiles .map( f => ` ### ${f.filename} \`\`\` ${f.content.slice(0, 300)}${f.content.length > 300 ? '\n... (truncated)' : ''} \`\`\` ` ) .join('\n') : '- No config files found' } ## Script Files ${ projectStructure.scriptFiles.length > 0 ? projectStructure.scriptFiles .map( f => ` ### ${f.filename} \`\`\` ${f.content.slice(0, 400)}${f.content.length > 400 ? '\n... (truncated)' : ''} \`\`\` ` ) .join('\n') : '- No script files found' } ## Existing Patterns Context ${ existingPatterns ? ` ### Current Patterns (${existingPatterns.length}) ${existingPatterns .map( (pattern, index) => ` #### ${index + 1}. ${pattern} ` ) .join('')} ` : 'No existing patterns provided.' } ## Pattern Generation Requirements 1. **Analyze project-specific content types** that need masking based on actual file content 2. **Identify sensitive data patterns** in code and documentation shown above 3. **Generate regex patterns** for consistent content masking 4. **Create appropriate replacements** that maintain context 5. **Ensure patterns don't conflict** with existing ones 6. **Provide clear descriptions** for each pattern ## Required Output Format Please provide custom pattern configuration in JSON format: \`\`\`json { "patterns": [ { "name": "pattern-name", "pattern": "regex-pattern", "replacement": "replacement-text", "description": "pattern-description", "category": "pattern-category" } ], "recommendations": ["list", "of", "recommendations"], "conflicts": ["any", "potential", "conflicts"] } \`\`\` `; const instructions = ` # Custom Pattern Configuration Instructions This analysis provides **actual project file contents** for comprehensive pattern generation. ## Analysis Scope - **Project Path**: ${projectPath} - **Package Files**: ${projectStructure.packageFiles.length} found - **Environment Files**: ${projectStructure.environmentFiles.length} found - **Config Files**: ${projectStructure.configFiles.length} found - **Script Files**: ${projectStructure.scriptFiles.length} found - **Total Files Analyzed**: ${projectStructure.totalFiles} - **Existing Patterns**: ${existingPatterns?.length || 0} patterns ## Next Steps 1. **Submit the configuration prompt** to an AI agent for pattern analysis 2. **Parse the JSON response** to get custom patterns and recommendations 3. **Review generated patterns** for accuracy and completeness 4. **Implement patterns** in the content masking system ## Expected AI Response Format The AI will return a JSON object with: - \`patterns\`: Array of custom pattern configurations - \`recommendations\`: Best practices and implementation guidance - \`conflicts\`: Potential conflicts with existing patterns ## Usage Example \`\`\`typescript const result = await configureCustomPatterns({ projectPath, existingPatterns }); // Submit result.configurationPrompt to AI agent // Parse AI response for custom pattern configuration \`\`\` `; const result = { configurationPrompt: customPatternPrompt, instructions, actualData: { projectStructure, summary: { totalFiles: projectStructure.totalFiles, packageFiles: projectStructure.packageFiles.length, environmentFiles: projectStructure.environmentFiles.length, configFiles: projectStructure.configFiles.length, scriptFiles: projectStructure.scriptFiles.length, }, }, }; return { content: [ { type: 'text', text: `# Custom Pattern Configuration\n\n${result.instructions}\n\n## AI Configuration Prompt\n\n${result.configurationPrompt}`, }, ], }; } catch (error) { throw new McpAdrError( `Failed to configure custom patterns: ${error instanceof Error ? error.message : String(error)}`, 'CONFIGURATION_ERROR' ); } }
• src/types/tool-arguments.ts:240-243 (schema)

  TypeScript interface defining the expected input arguments for the configure_custom_patterns tool.
  export interface ConfigureCustomPatternsArgs { projectPath: string; existingPatterns?: string[]; }
• src/utils/server-context-generator.ts:177-180 (registration)

  The tool name 'configure_custom_patterns' is registered/listed in the server context generator's comprehensive tool catalog, confirming its inclusion in the MCP toolset.
  name: 'configure_custom_patterns', description: 'Configure custom security patterns for content analysis', }, {