Flow Metrics Coach

Analyze flow metrics for project $0.

Data collection

1. Use ticket_list to get all tickets (ticket_list returns 100 tickets per call: page with offset until you have total of them)
2. Use report_times for each ticket's lead and cycle time in whole days
3. Use event_log for the state transitions themselves: its tag_added and tag_removed events with prefix state carry their timestamps, which stage durations and weekly throughput need. Read it oldest first: start with after: 0, then pass the sequence of the last event received as after, until a page has fewer events than limit (default 50). Without after, event_log returns the newest events first, and paging from there skips everything older. ticket_history doesn't help here: it lists score and title changes, not tag changes

Metrics to calculate

Lead Time

Time from ticket creation (or state:backlog assignment) to state:done.

Cycle Time

Time from state:wip to state:done (active work only).

Throughput

Number of tickets reaching state:done per week.

Stage Duration

Average time spent in each state: tag value.

Output

Summary statistics

• Mean, median, and 85th percentile for lead time and cycle time
• Weekly throughput trend

Stage breakdown

Bottleneck analysis

Identify the stage with the longest average duration and suggest:

• WIP limits
• Pairing or swarming
• Process changes

Recommendations

Provide 2-3 actionable suggestions based on the data, e.g.:

• "Tickets spend 4.2 days average in state:review. Consider pairing to reduce this."
• "Throughput dropped 30% last week. 5 tickets are stuck in state:wip."