内容指标

1. 总词数

   • 所有转录文本的总词数
   • 指示会话深度/参与度

2. 每会话平均词数

   • 总词数 / 会话数量

3. 发言者

   • 识别唯一的发言者（如果可用）
   • 追踪多发言者会话

4. 主题/关键词

   • 从转录文本中提取常见主题
   • 识别教练主题

分析脚本

analyze_fieldy.py

分析 Fieldy 转录数据并计算指标。

用法：

python scripts/analyze_fieldy.py input.json --output analysis.json

输出：

{
  "date": "2026-01-04",
  "summary": {
    "total_sessions": 12,
    "total_duration_minutes": 145.3,
    "avg_session_minutes": 12.1,
    "total_words": 5234,
    "avg_words_per_session": 436,
    "first_session": "2026-01-04T08:15:00",
    "last_session": "2026-01-04T17:45:00"
  },
  "sessions": [
    {
      "timestamp": "2026-01-04T08:15:00",
      "duration_minutes": 8.5,
      "word_count": 342,
      "segment_count": 15
    }
  ]
}

generate_fieldy_summary.py

为每日邮件报告生成格式化摘要。

用法：

python scripts/generate_fieldy_summary.py input.json \
    --output summary.json \
    --format report

输出（报告格式）：

{
  "title": "Fieldy 每日摘要",
  "subtitle": "2026年1月4日，星期五",
  "metrics": [
    {
      "label": "总会话数",
      "value": "12",
      "trend": "up",
      "change": "+2"
    },
    {
      "label": "总时长",
      "value": "145 分钟",
      "trend": "up",
      "change": "+15 分钟"
    },
    {
      "label": "平均会话长度",
      "value": "12.1 分钟",
      "trend": "neutral"
    }
  ],
  "sections": [
    {
      "title": "会话时间线",
      "type": "timeline",
      "data": {
        "events": [...]
      }
    }
  ]
}

extract_insights.py

使用文本分析从转录文本中提取关键主题和见解。

用法：

python scripts/extract_insights.py input.json --top-keywords 10

功能：

• 关键词频率分析
• 主题聚类
• 情感指标
• 常见短语

报告模板

每日邮件摘要

每日 Fieldy 邮件的标准格式：

主题： Fieldy 每日摘要 - [日期]

内容：

• 总会话数
• 总时长（分钟）
• 平均会话长度
• 会话时间线（首次/末次会话时间）
• 显著关键词或主题（如果可用）

每周汇总

一周的聚合指标：

• 每日会话数
• 趋势分析（增加/减少）
• 最繁忙的一天
• 每周总时长

与智能体的集成

Fieldy 智能体工作流

# 1. 读取今日的 Fieldy 数据
fieldy_file = f"data/fieldy/fieldy_{today}.json"

# 2. 使用 fieldy-analysis 技能进行分析
python scripts/analyze_fieldy.py {fieldy_file} --output analysis.json

# 3. 为报告生成摘要
python scripts/generate_fieldy_summary.py analysis.json --format report

# 4. 使用 report-generation 技能创建 HTML
python ../report-generation/scripts/generate_html_report.py \
    summary.json --template daily-summary --output fieldy_report.html

# 5. 委托 communication-agent 发送邮件
task(agent="communication-agent",
     prompt=f"Send fieldy_report.html to dave+mega@flycowgames.com")

报告智能体集成

用于结合 Fieldy 与其他指标的每周报告：

# 聚合一周的 Fieldy 数据
fieldy_weekly = aggregate_fieldy_week(start_date, end_date)

# 与 GitHub、Skillz 等数据结合
combined_report = {
    "fieldy": fieldy_weekly,
    "github": github_weekly,
    "skillz": skillz_weekly
}

# 生成全面的每周报告
python ../report-generation/scripts/generate_html_report.py \
    combined_report.json --template weekly-summary

数据质量

验证检查

1. 数据缺失

   • 检查空转录
   • 验证时间戳格式
   • 确保时长计算为正数

2. 异常值

   • 标记 > 60 分钟的会话（可能是错误）
   • 标记 < 1 分钟的会话（不完整）
   • 检查重复的时间戳

3. 完整性

   • 验证所有必填字段是否存在
   • 检查 transcriptions 数组非空
   • 验证 metadata 结构

调度

Fieldy 智能体每日在以下时间运行：

• 时间： 11:00 AM UTC (3:00 AM PT)
• 触发器： systemd 定时器 (daily-fieldy-email.timer)
• 收件人： dave+mega@flycowgames.com

提示

1. 时长计算

   • 优先使用片段的开始/结束时间（如果可用）
   • 回退到词数估算（150 词/分钟）
   • 优雅处理缺失的时长数据

2. 时区处理

   • Fieldy 时间戳为 UTC
   • 为邮件显示转换为 PT
   • 确保日期边界正确

3. 错误处理

   • 优雅处理缺失的文件
   • 如果一个会话失败，继续处理其他会话
   • 记录错误以便调试

4. 性能

   • 缓存解析后的数据以避免重复解析
   • 如果会话很多，并行处理
   • 将关键词提取限制在前 N 个

每日报告示例

Fieldy 每日摘要
2026年1月4日，星期五

摘要：
✓ 12 次教练会话
✓ 总时长 145 分钟
✓ 平均会话长度 12.1 分钟

时间线：
• 首次会话：上午 8:15 PT
• 末次会话：下午 5:45 PT
• 最活跃时段：下午 2:00 - 下午 4:00

热门关键词：
• technique (提及 45 次)
• training (提及 38 次)
• form (提及 32 次)
• practice (提及 28 次)

参考

• Fieldy 数据位置：data/fieldy/fieldy_YYYY-MM-DD.json
• 归档位置：data/fieldy/archive/ (可选)
• 报告收件人：参见 config/email_accounts.json

每周安装数

1

仓库

dglowacki/mega-agent-v2

首次出现

1 天前

安全审计

Gen Agent Trust HubPassSocketPassSnykPass

安装于

zencoder1

amp1

cline1

openclaw1

opencode1

cursor1

1. Total Sessions

   • Count of transcription entries for the day
   • Each entry = one coaching session/call

2. Total Duration

   • Calculate from start/end times of transcription segments
   • Sum of all segment durations

3. Average Session Length

   • Total duration / number of sessions
   • Measured in minutes

4. Session Timeline

   • Timestamps of each session
   • Distribution throughout the day

Content Metrics

1. Total Words

   • Word count across all transcriptions
   • Indicates session depth/engagement

2. Average Words per Session

   • Total words / number of sessions

3. Speakers

   • Identify unique speakers (if available)
   • Track multi-speaker sessions

4. Topics/Keywords

   • Extract common topics from transcriptions
   • Identify coaching themes

Analysis Scripts

analyze_fieldy.py

Analyze Fieldy transcription data and calculate metrics.

Usage:

python scripts/analyze_fieldy.py input.json --output analysis.json

Output:

{
  "date": "2026-01-04",
  "summary": {
    "total_sessions": 12,
    "total_duration_minutes": 145.3,
    "avg_session_minutes": 12.1,
    "total_words": 5234,
    "avg_words_per_session": 436,
    "first_session": "2026-01-04T08:15:00",
    "last_session": "2026-01-04T17:45:00"
  },
  "sessions": [
    {
      "timestamp": "2026-01-04T08:15:00",
      "duration_minutes": 8.5,
      "word_count": 342,
      "segment_count": 15
    }
  ]
}

generate_fieldy_summary.py

Generate formatted summary for daily email report.

Usage:

python scripts/generate_fieldy_summary.py input.json \
    --output summary.json \
    --format report

Output (report format):

{
  "title": "Fieldy Daily Summary",
  "subtitle": "Friday, January 4, 2026",
  "metrics": [
    {
      "label": "Total Sessions",
      "value": "12",
      "trend": "up",
      "change": "+2"
    },
    {
      "label": "Total Duration",
      "value": "145 min",
      "trend": "up",
      "change": "+15 min"
    },
    {
      "label": "Avg Session Length",
      "value": "12.1 min",
      "trend": "neutral"
    }
  ],
  "sections": [
    {
      "title": "Session Timeline",
      "type": "timeline",
      "data": {
        "events": [...]
      }
    }
  ]
}

extract_insights.py

Extract key topics and insights from transcriptions using text analysis.

Usage:

python scripts/extract_insights.py input.json --top-keywords 10

Features:

• Keyword frequency analysis
• Topic clustering
• Sentiment indicators
• Common phrases

Report Templates

Daily Email Summary

Standard format for daily Fieldy email:

Subject: Fieldy Daily Summary - [Date]

Content:

• Total sessions count
• Total duration in minutes
• Average session length
• Session timeline (first/last session times)
• Notable keywords or topics (if available)

Weekly Rollup

Aggregate metrics for the week:

• Sessions per day
• Trend analysis (increasing/decreasing)
• Busiest day
• Total weekly duration

Integration with Agents

Fieldy Agent Workflow

# 1. Read today's Fieldy data
fieldy_file = f"data/fieldy/fieldy_{today}.json"

# 2. Analyze with fieldy-analysis skill
python scripts/analyze_fieldy.py {fieldy_file} --output analysis.json

# 3. Generate summary for report
python scripts/generate_fieldy_summary.py analysis.json --format report

# 4. Use report-generation skill to create HTML
python ../report-generation/scripts/generate_html_report.py \
    summary.json --template daily-summary --output fieldy_report.html

# 5. Delegate to communication-agent to send email
task(agent="communication-agent",
     prompt=f"Send fieldy_report.html to dave+mega@flycowgames.com")

Reporting Agent Integration

For weekly reports combining Fieldy with other metrics:

# Aggregate week's Fieldy data
fieldy_weekly = aggregate_fieldy_week(start_date, end_date)

# Combine with GitHub, Skillz, etc.
combined_report = {
    "fieldy": fieldy_weekly,
    "github": github_weekly,
    "skillz": skillz_weekly
}

# Generate comprehensive weekly report
python ../report-generation/scripts/generate_html_report.py \
    combined_report.json --template weekly-summary

Data Quality

Validation Checks

1. Missing Data

   • Check for empty transcriptions
   • Validate timestamp format
   • Ensure duration calculations are positive

2. Outliers

   • Flag sessions > 60 minutes (likely errors)
   • Flag sessions < 1 minute (incomplete)
   • Check for duplicate timestamps

3. Completeness

   • Verify all required fields present
   • Check transcriptions array not empty
   • Validate metadata structure

Scheduling

The Fieldy Agent runs daily at:

• Time: 11:00 AM UTC (3:00 AM PT)
• Trigger: systemd timer (daily-fieldy-email.timer)
• Recipients: dave+mega@flycowgames.com

Tips

1. Duration Calculation

   • Use segment start/end times when available
   • Fall back to word count estimation (150 words/min)
   • Handle missing duration gracefully

2. Timezone Handling

   • Fieldy timestamps are in UTC
   • Convert to PT for email display
   • Ensure date boundaries are correct

3. Error Handling

   • Handle missing files gracefully
   • Continue processing if one session fails
   • Log errors for debugging

4. Performance

   • Cache parsed data to avoid re-parsing
   • Process sessions in parallel if many
   • Limit keyword extraction to top N

Example Daily Report

Fieldy Daily Summary
Friday, January 4, 2026

Summary:
✓ 12 total coaching sessions
✓ 145 minutes total duration
✓ 12.1 minutes average session length

Timeline:
• First session: 8:15 AM PT
• Last session: 5:45 PM PT
• Most active period: 2:00 PM - 4:00 PM

Top Keywords:
• technique (45 mentions)
• training (38 mentions)
• form (32 mentions)
• practice (28 mentions)

Reference

• Fieldy data location: data/fieldy/fieldy_YYYY-MM-DD.json
• Archive location: data/fieldy/archive/ (optional)
• Report recipients: See config/email_accounts.json

Weekly Installs

1

Repository

dglowacki/mega-agent-v2

First Seen

1 day ago

Security Audits

Gen Agent Trust HubPassSocketPassSnykPass

Installed on

zencoder1

amp1

cline1

openclaw1

opencode1

cursor1