custom_gates

Custom Gates Examples

Advanced quality gate patterns — dynamic thresholds, composite scoring, and metric history tracking.

Files

• dynamic_thresholds.py — Adjust gate thresholds based on git change size
• composite_gates.py — Weighted composite score from multiple metrics
• metric_history.py — Track metrics over time and detect regressions
• pyqual.yaml — Pipeline config for custom gates

dynamic_thresholds.py

Adjusts coverage thresholds based on the number of changed Python files:

• Large changes (>10 files): lower threshold (60%)
• Small changes (≤10 files): higher threshold (80%)

python dynamic_thresholds.py

composite_gates.py

Combines coverage, complexity, lint errors, and security issues into a single weighted score (0–100). Gates pass only if both individual gates AND the composite score meet thresholds.

python composite_gates.py

Weights:

Component	Weight	Source
Coverage	35%	pytest-cov
Complexity	25%	code2llm
Lint errors	20%	ruff
Security	20%	bandit

metric_history.py

Stores metric snapshots in .pyqual/metric_history.json and detects regressions between runs. Fails if any metric regresses beyond a configurable tolerance.

python metric_history.py

Features:

• Timestamped metric snapshots
• Per-metric trend analysis (improving / degrading / stable)
• Configurable regression tolerance (default: 2%)
• Distinguishes "higher is better" vs "lower is better" metrics

Usage

cd examples/custom_gates

# Run individual examples
python dynamic_thresholds.py
python composite_gates.py
python metric_history.py

# Run the pipeline
pyqual run -c pyqual.yaml

Key Concepts

• GateConfig — defines a single metric + operator + threshold
• GateSet — collection of gates with metric collection from .pyqual/ artifacts
• Gate.check(metrics) — evaluates one gate against collected metrics
• gate_set.check_all(workdir) — collects metrics and checks all gates
• gate_set._collect_metrics(workdir) — gathers metrics from all sources

pyqual.yaml

See pyqual.yaml for the example pipeline configuration.