sdlc-metrics

SDLC Metrics

Flow metrics and analysis for the Infiquetra Mount Olympus Kanban board using GitHub timeline events. Measures cycle time, throughput, WIP age, and flow efficiency — with targets by issue type.

Script Location

$INFIQUETRA_SDLC_PATH/../infiquetra-claude-plugins/plugins/sdlc-manager/scripts/sdlc_manager.py

If $INFIQUETRA_SDLC_PATH is unset, use ~/workspace/infiquetra/infiquetra-sdlc as the default base path.

IMPORTANT: Always use python3 (not python) to run the script.

How Timeline Metrics Work

The script queries timelineItems on GitHub issues to find ProjectV2ItemFieldValueEvent entries showing status column changes. From these events it calculates:

• Time in each column: Timestamp of entry into column -> timestamp of exit from column
• Cycle time: Time from first entry into "In Development" through "Deployed"
• Percentiles: P50 (median), P85 (target benchmark), P95 (worst-case tail)
• WIP age: For in-progress items, elapsed time since entering the current column

Core Operations

Cycle Time

# Cycle time for all issue types over last 30 days
python3 sdlc_manager.py metrics cycle-time --project mount-olympus

# Filter by days window
python3 sdlc_manager.py metrics cycle-time --project mount-olympus --days 14

# Filter to specific issue type
python3 sdlc_manager.py metrics cycle-time --project mount-olympus --type capability
python3 sdlc_manager.py metrics cycle-time --project mount-olympus --type enhancement
python3 sdlc_manager.py metrics cycle-time --project mount-olympus --type defect

Sample Output

Cycle Time Report — mount-olympus (last 30 days)

              P50      P85      P95    Target   Status
capability    3.2d     4.8d     7.1d   < 5d     ✓ On target
enhancement   1.1d     1.8d     3.2d   < 2d     ✓ On target
defect        0.4d     0.9d     1.5d   < 1d     ✓ On target
exploration   1.5d     2.6d     3.0d   < 3d     ✓ On target

Items measured: 18 (9 capability, 5 enhancement, 3 defect, 1 exploration)
Slowest item:  #38 capability "User onboarding service" — 7.1 days
               Bottleneck: E2E Testing (2.3 days)

Throughput

# Throughput over last 4 weeks (default)
python3 sdlc_manager.py metrics throughput --project mount-olympus

# Specify weeks window
python3 sdlc_manager.py metrics throughput --project mount-olympus --weeks 8

# Filter by issue type
python3 sdlc_manager.py metrics throughput --project mount-olympus --type capability

WIP Age

# Current WIP age distribution (all in-progress items)
python3 sdlc_manager.py metrics wip-age --project mount-olympus

Output shows each in-progress item, which column it is in, and how long it has been there. Items exceeding age thresholds are flagged.

Per-Column Time Breakdown

# How long a specific issue spent in each column
python3 sdlc_manager.py metrics column-time --project mount-olympus --number 42

Useful for diagnosing where an individual item got stuck.

Targets by Issue Type

Cycle Time Targets (P85)

Issue Type	P85 Target	Notes
Capability	< 5 days	End-to-end: In Development -> Deployed
Enhancement	< 2 days	Smaller scope
Defect (general)	< 1 day	Highest urgency
Exploration	< 3 days	Timeboxed research
Context Update	< 1 day	Documentation

Defect SLA Targets

Priority	SLA	Description
Critical	4 hours	System down, data loss, security breach
High	1 day	Major functionality broken
Medium	3 days	Minor functionality broken
Low	When capacity	Cosmetic or rare edge case

Throughput Targets (per week)

Issue Type	Target	Direction
Capabilities	2-4	Higher is better
Enhancements	5-10	Higher is better
Defects	< 2	Lower is better

WIP Age Thresholds (flag as aging)

Column	Age Threshold
Ready	> 2 days
In Development	> 5 days
E2E Testing	> 1 day
Deployment Ready	> 2 days

Flow Efficiency Target

Target: > 50%

Formula: (time in In Development + time in E2E Testing) / total cycle time

Active columns are In Development and E2E Testing. Waiting columns are Ready and Deployment Ready. High wait time in Deployment Ready typically indicates deployment window constraints.

Natural Language Examples

"What's our cycle time for capabilities this month?" -> metrics cycle-time --project mount-olympus --type capability --days 30

"How many capabilities did we ship last week?" -> metrics throughput --project mount-olympus --weeks 1 --type capability

"How old are our in-progress items?" -> metrics wip-age --project mount-olympus

"How long has issue #42 been in development?" -> metrics column-time --project mount-olympus --number 42

"Are we meeting our targets?" -> Run metrics cycle-time and metrics throughput, compare to targets above

"Are any defects violating SLA?" -> metrics wip-age --project mount-olympus — look for defects flagged as aging

"Compare this month to last month" -> metrics cycle-time --project mount-olympus --days 30 vs. --days 60 (compare first/second half)

Interpreting Results

When Cycle Time is Over Target

1. Run column-time on slow items to find where they got stuck
2. Check E2E Testing aging — manual testing is often the bottleneck
3. Check Deployment Ready aging — deployment window delays inflate cycle time
4. Look for blocked items — blocked label can stall flow for days
5. Check WIP in In Development — high WIP creates context switching, slows everything

When Throughput is Low

1. Check WIP in In Development — if near or over limit, team may be stuck
2. Check In Development aging — items not completing drive low throughput
3. Check Ready column — if empty, team may lack prioritized work
4. Look at Objective completion — are all capabilities tied to a blocked Objective?

When WIP Age is High

1. Identify the specific items and ask about blockers in standup
2. Consider swarming — additional developers on a stuck item can unblock it
3. Check for external blockers — dependencies on other teams or systems
4. Escalate per policy: blocked > 4 hours -> tech lead; blocked > 1 day -> EM

When Flow Efficiency is Low (< 50%)

1. High wait in Deployment Ready -> deployment scheduling delays
2. High wait in Ready -> context switching or team not pulling promptly
3. Multiple handoffs -> check if items are bouncing between columns
4. Blocked time -> blocked items count as wait time, reducing efficiency

Workflow Examples

Weekly Metrics Review

# Run for mount-olympus board
python3 sdlc_manager.py metrics cycle-time --project mount-olympus --days 7
python3 sdlc_manager.py metrics throughput --project mount-olympus --weeks 4
python3 sdlc_manager.py metrics wip-age --project mount-olympus

Compare each metric to the targets table. Share summary in #mount-olympus Discord channel.

SLA Check for Defects

# Get all in-progress items (includes defects with their age)
python3 sdlc_manager.py metrics wip-age --project mount-olympus

# For any defect flagged as aging, check full timeline
python3 sdlc_manager.py metrics column-time --project mount-olympus --number <N>

Cross-reference priority label (critical/high-priority/medium-priority) against SLA targets above.

Diagnosing a Bottleneck

# 1. Check which column has the most aging items
python3 sdlc_manager.py metrics wip-age --project mount-olympus

# 2. Check cycle time breakdown to see which column adds the most time
python3 sdlc_manager.py metrics cycle-time --project mount-olympus --days 30

# 3. Drill into specific slow items
python3 sdlc_manager.py metrics column-time --project mount-olympus --number <N>

Reference Documents

• references/metrics-targets.md — Complete targets, definitions, and interpretation guide
• skills/sdlc-board/references/kanban-workflow.md — Column definitions and WIP limits