Scan signals
Start broad across lead time, failure rate and review load.
Progressive DiscoveryQuality Engineering Demo
Engineering Metrics Agent
A metrics agent that scans Git, CI and Jira signals, retrieves metric definitions, and writes a source-backed trend report.
Structure Blueprint
Clickable pattern simulator
Interactive Browser Demo
This simulator does not replace the CLI run, but it makes each step, state, log, gate and final output visible in the browser.
Live pipeline
State inspector
Execution log
Current output
Step-by-step Design
Focus anomalies
Drill into metrics whose slope changed.
Progressive DiscoveryRetrieve definitions
Bind each metric to the team's agreed definition.
RAG PipelineBuild report
Generate trend, interpretation and likely cause.
Prompt ChainingCheck governance
Flag metrics that could be misused or lack source coverage.
Approval GateExpose trace
Show source rows, definition, slope and confidence.
Observability HarnessScreenshots
Pattern Mapping
| Pattern | Design role |
|---|---|
| Progressive Discovery | Do not load every metric; forage, focus, then drill. |
| RAG Pipeline | Metric definitions need source-backed retrieval. |
| Observability Harness | Trend, source coverage and confidence are part of the product. |
| Approval Gate | Potentially misleading metrics should be flagged before publication. |
Design Insights
Metrics need definitions
The agent should explain what a metric means before explaining why it moved.
Slope beats snapshot
A week-over-week slope catches quality drift earlier than a static threshold.
Governance prevents metric theatre
A report should warn when a metric can be gamed or lacks source coverage.
Run Path
cd workshops/workshop4-engineering-metrics
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
python src/main.py
open visual/index.htmlworkshops/workshop4-engineering-metrics/README.mdimages/training/visual/workshop4-engineering-metrics-blueprint.png