M.I.M.I.R - Multi-agent Intelligent Memory & Insight Repository

interface ReportData { agent: string; benchmark: string; model: string; result: { output: string; conversationHistory: Array<{ role: string; content: string }>; tokens: { input: number; output: number }; toolCalls?: number; intermediateSteps?: any[]; }; scores: { categories: Record<string, number>; total: number; feedback: Record<string, string>; }; } /** * Generate a comprehensive validation report for agent execution * * Creates a markdown-formatted report containing: * - Execution metadata (agent, benchmark, model, date) * - Performance metrics (tokens, tool calls) * - Scoring breakdown by category with feedback * - Complete agent output * - Full conversation history * * This report is used for: * - Agent validation and benchmarking * - Performance analysis and optimization * - Debugging agent behavior * - Documentation and audit trails * * @param data - Report data including agent info, execution results, and scores * @returns Markdown-formatted report string * * @example * ```ts * const reportData = { * agent: 'worker-a3f2b8c1', * benchmark: 'golang-crypto-task', * model: 'gpt-4.1', * result: { * output: 'Task completed successfully...', * conversationHistory: [ * { role: 'user', content: 'Implement RSA encryption' }, * { role: 'assistant', content: 'I will implement...' } * ], * tokens: { input: 1500, output: 2000 }, * toolCalls: 12 * }, * scores: { * categories: { * 'Correctness': 25, * 'Code Quality': 20, * 'Documentation': 15 * }, * total: 85, * feedback: { * 'Correctness': 'Implementation is correct and handles edge cases', * 'Code Quality': 'Well-structured with good error handling', * 'Documentation': 'Clear comments and examples provided' * } * } * }; * * const report = generateReport(reportData); * fs.writeFileSync('validation-report.md', report); * ``` */ export function generateReport(data: ReportData): string { return ` # Agent Validation Report **Agent**: ${data.agent} **Benchmark**: ${data.benchmark} **Model**: ${data.model} **Date**: ${new Date().toISOString().split('T')[0]} **Total Score**: ${data.scores.total}/100 --- ## Execution Summary - **Tool Calls**: ${data.result.toolCalls || 0} - **Input Tokens**: ${data.result.tokens.input} - **Output Tokens**: ${data.result.tokens.output} - **Total Tokens**: ${data.result.tokens.input + data.result.tokens.output} --- ## Scoring Breakdown ${Object.entries(data.scores.categories) .map( ([category, score]) => ` ### ${category}: ${score} points **Feedback**: ${data.scores.feedback[category]} ` ) .join('\n')} --- ## Agent Output \`\`\` ${data.result.output} \`\`\` --- ## Conversation History ${data.result.conversationHistory .map( (msg) => ` ### ${msg.role.toUpperCase()} \`\`\` ${msg.content} \`\`\` ` ) .join('\n')} `.trim(); }