Reporting Skill -- Daily Reports, Retrospectives, and Dashboard

Purpose

Generate structured reports for human consumption:

1. •Daily report: Published every day at 17:00 to #em-dashboard. System health snapshot.
2. •Weekly retrospective: Published every Friday at 17:30 to #em-dashboard. Conversion funnel, WoW comparisons, findings, Kaizen log.
3. •Dashboard formatter: Formats data for Slack markdown posting (agent status, metrics, alerts).

Daily Report

Trigger: 17:00 every day

Components:

1. •

   Executive summary (3-4 bullet points)

   • •Applications sent today
   • •Response rate (cumulative)
   • •Interviews booked today
   • •Inbound leads received
2. •

   Agent performance (per-agent section)

   • •Jay: Jobs researched, avg confidence, staleness, end-client deduction
   • •Z: CRM update latency, dupe rate, Hot List publication time, completeness
   • •Rick: Matching cycle time, avg match score, trifecta pass rate, inbound response
   • •Leroy: Apps executed, profile health, inbound detection latency, execution errors
3. •

   Alerts section

   • •Any alerts from the day (critical/high only)
4. •

   System health scorecard

   • •Agent uptime (% ACTIVE/IDLE combined)
   • •Data quality score
   • •Error rate
   • •Message volume
5. •

   Kaizen observation of the day

   • •Category, observation, proposed action

Format: Slack markdown with emojis and status bars

Weekly Retrospective

Trigger: Friday 17:30

Components:

1. •

   Conversion funnel (WoW)

   • •Jobs researched (this week vs. last week)
   • •Matches generated
   • •Applications sent
   • •Responses received
   • •Interviews booked
   • •Offers received
   • •Conversion rates at each stage
2. •

   Per-agent performance (WoW)

   • •Each agent's key metrics compared to last week
   • •Trends (up/down arrows)
   • •Notable events
3. •

   Top events of the week

   • •Profile bans, visa urgencies, inbound surges
   • •System incidents
4. •

   Retrospective findings

   • •What worked well? What didn't?
   • •Blockers encountered
   • •Dependencies that caused delays
5. •

   Proposed adjustments (awaiting human approval)

   • •Weight changes
   • •Quota adjustments
   • •Process improvements
6. •

   Kaizen digest for the week

   • •All daily observations compiled
   • •Categorized by type (PROCESS, QUALITY, SPEED, COST, RISK)

Dashboard Formatter

Formats system data for Slack markdown. Used by daily_report.py and weekly_retrospective.py.

Templates:

• •Agent status bars: Shows ACTIVE/IDLE/BUSY/SLOW/ERROR/DEAD with visual indicator
• •Metric scorecards: Target vs. actual with status (🟢 MET / 🟡 AT RISK / 🔴 MISS)
• •Alert summary: Icon + severity + title + action
• •Pipeline visualization: Numbers flowing through system stages

Implementation

• •daily_report.py: Runs at 17:00. Aggregates data from quotas, alerts, heartbeats. Formats and posts to Slack.
• •weekly_retrospective.py: Runs Friday 17:30. Analyzes 7-day history. Compiles findings and Kaizen.
• •dashboard_formatter.py: Helper functions for Slack markdown formatting.

All reports include:

• •Timestamp
• •Data freshness (when was this last calculated?)
• •Confidence (based on data quality)
• •Link to detailed logs for human investigation