Adoption-proof examples¶

These examples show realistic, artifact-first usage that maps to day-to-day PR and release review flow.

Canonical artifact-producing run:

python -m sdetkit gate fast --format json --stable-json --out build/gate-fast.json
python -m sdetkit gate release --format json --out build/release-preflight.json
python -m sdetkit doctor

Example 1: Team runs `gate fast`, `gate release`, `doctor`¶

python -m sdetkit gate fast --format json --stable-json --out build/gate-fast.json
python -m sdetkit gate release --format json --out build/release-preflight.json
python -m sdetkit doctor

Why this sequence works: - gate fast gives immediate pre-merge signal. - gate release gives release-level aggregate decision input. - doctor validates runtime/setup assumptions that can explain unexpected failures.

Example 2: Artifact-first triage path¶

Use this deterministic order: 1. Read build/release-preflight.json. 2. If failed_steps includes gate_fast, read build/gate-fast.json. 3. Only then inspect raw logs for low-level details.

This keeps triage focused on explicit machine-readable decision points.

Example 3: What blocked¶

A practical blocked state looks like: - release-preflight.json has ok: false - failed_steps is non-empty - failure points to the next deterministic remediation target

At this stage, the team has enough structure to decide “not ready yet” without narrative guesswork.

Example 4: What to fix next¶

Teams can choose the next action from the first failing artifact field instead of broad, unfocused investigation.

Typical next-step pattern: - fix the earliest failing gate condition - re-run gate fast - re-run gate release - keep doctor clean before final review

Example 5: How this helps real review/release flow¶

In practical review operations: - engineers attach artifact outputs in PR context - reviewers validate ok / failed_steps quickly - release owners use the same outputs for go/no-go calls

This creates a single evidence thread from implementation review through release decision, without changing core repo workflows.