usage_stats
View usage statistics for multiple AI models including call count, success rate, and average response time.
Instructions
查看各模型的使用统计（调用次数、成功率、平均耗时）
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Implementation Reference
- src/server.ts:583-592 (handler)The case handler for 'usage_stats' tool. It calls globalStats.formatStats() and returns the formatted statistics as a text response.
case 'usage_stats': { return { content: [ { type: 'text', text: globalStats.formatStats(), }, ], }; } - src/server.ts:259-266 (schema)Tool registration definition for 'usage_stats' with name, description, and empty input schema (no parameters).
{ name: 'usage_stats', description: '查看各模型的使用统计（调用次数、成功率、平均耗时）', inputSchema: { type: 'object', properties: {}, }, }, - src/server.ts:259-267 (registration)The 'usage_stats' tool is registered as part of the tools array (presumably in a ListTools handler) alongside other tools like history_context and team_dashboard.
{ name: 'usage_stats', description: '查看各模型的使用统计（调用次数、成功率、平均耗时）', inputSchema: { type: 'object', properties: {}, }, }, { - src/collaboration/stats.ts:73-218 (helper)The UsageStats class that provides all statistics data: recording calls, computing per-model and global stats, formatting output. The singleton instance 'globalStats' is exported and used by the handler.
export class UsageStats { private records: CallRecord[] = []; private readonly maxRecords: number; private readonly startTime: number; constructor(maxRecords: number = 1000) { this.maxRecords = maxRecords; this.startTime = Date.now(); } /** * 记录一次调用 */ record(record: Omit<CallRecord, 'duration'>): void { const fullRecord: CallRecord = { ...record, duration: record.endTime - record.startTime, }; this.records.push(fullRecord); // 限制记录数量 if (this.records.length > this.maxRecords) { this.records = this.records.slice(-this.maxRecords); } } /** * 开始计时 * @returns 结束计时的函数，调用时传入 success 和可选的 error */ startTimer(model: string, taskType?: string): (success?: boolean, error?: string) => void { const startTime = Date.now(); return (success: boolean = true, error?: string) => { this.record({ model, startTime, endTime: Date.now(), success, error, taskType, }); }; } /** * 获取模型统计 */ getModelStats(model: string): ModelStats | null { const modelRecords = this.records.filter(r => r.model === model); if (modelRecords.length === 0) return null; const successRecords = modelRecords.filter(r => r.success); const durations = successRecords.map(r => r.duration); return { model, totalCalls: modelRecords.length, successCalls: successRecords.length, failedCalls: modelRecords.length - successRecords.length, successRate: successRecords.length / modelRecords.length, avgDuration: durations.length > 0 ? durations.reduce((a, b) => a + b, 0) / durations.length : 0, minDuration: durations.length > 0 ? Math.min(...durations) : 0, maxDuration: durations.length > 0 ? Math.max(...durations) : 0, totalDuration: durations.reduce((a, b) => a + b, 0), }; } /** * 获取全局统计 */ getGlobalStats(): GlobalStats { const models = new Set(this.records.map(r => r.model)); const modelStats: ModelStats[] = []; for (const model of models) { const stats = this.getModelStats(model); if (stats) modelStats.push(stats); } const successRecords = this.records.filter(r => r.success); const totalDuration = this.records .filter(r => r.success) .reduce((a, b) => a + b.duration, 0); return { totalCalls: this.records.length, totalSuccess: successRecords.length, totalFailed: this.records.length - successRecords.length, totalDuration, avgDuration: successRecords.length > 0 ? totalDuration / successRecords.length : 0, models: modelStats.sort((a, b) => b.totalCalls - a.totalCalls), since: this.startTime, }; } /** * 格式化统计信息 */ formatStats(): string { const stats = this.getGlobalStats(); const uptime = Math.floor((Date.now() - stats.since) / 1000 / 60); let output = `📊 **使用统计** (运行 ${uptime} 分钟)\n\n`; output += `| 指标 | 数值 |\n|------|------|\n`; output += `| 总调用 | ${stats.totalCalls} 次 |\n`; output += `| 成功 | ${stats.totalSuccess} 次 |\n`; output += `| 失败 | ${stats.totalFailed} 次 |\n`; output += `| 平均耗时 | ${(stats.avgDuration / 1000).toFixed(2)}s |\n`; output += `| 总耗时 | ${(stats.totalDuration / 1000).toFixed(1)}s |\n\n`; if (stats.models.length > 0) { output += `### 各模型统计\n\n`; output += `| 模型 | 调用 | 成功率 | 平均耗时 |\n|------|------|--------|----------|\n`; for (const model of stats.models) { output += `| ${model.model} | ${model.totalCalls} | ${(model.successRate * 100).toFixed(0)}% | ${(model.avgDuration / 1000).toFixed(2)}s |\n`; } } return output; } /** * 清空统计 */ clear(): void { this.records = []; } /** * 获取最近的调用记录 */ getRecentRecords(count: number = 10): CallRecord[] { return this.records.slice(-count); } } /** 全局统计实例 */ export const globalStats = new UsageStats(); - src/collaboration/stats.ts:9-68 (schema)Type definitions (interfaces) used by the UsageStats system: CallRecord (individual call), ModelStats (per-model aggregation), and GlobalStats (overall summary).
export interface CallRecord { /** 模型名称 */ model: string; /** 开始时间 */ startTime: number; /** 结束时间 */ endTime: number; /** 耗时（毫秒） */ duration: number; /** 是否成功 */ success: boolean; /** 错误信息 */ error?: string; /** 任务类型 */ taskType?: string; } /** * 模型统计信息 */ export interface ModelStats { /** 模型名称 */ model: string; /** 总调用次数 */ totalCalls: number; /** 成功次数 */ successCalls: number; /** 失败次数 */ failedCalls: number; /** 成功率 */ successRate: number; /** 平均耗时（毫秒） */ avgDuration: number; /** 最短耗时 */ minDuration: number; /** 最长耗时 */ maxDuration: number; /** 总耗时 */ totalDuration: number; } /** * 全局统计信息 */ export interface GlobalStats { /** 总调用次数 */ totalCalls: number; /** 总成功次数 */ totalSuccess: number; /** 总失败次数 */ totalFailed: number; /** 总耗时 */ totalDuration: number; /** 平均耗时 */ avgDuration: number; /** 各模型统计 */ models: ModelStats[]; /** 统计开始时间 */ since: number; }