Previewing Analysis and Gates Without a New Scan

Overview

StretchDev separates two kinds of work: running a scan, which writes a new record to your history, and code intelligence, which computes and returns results without changing anything. The intelligence views are perfect for exploration and policy tuning because they are guaranteed read-only — no scan is written, nothing is mutated. There are three: analyze a stored scan, preview a quality gate, and score diff risk.

Analyze a stored scan

Use this to re-read the results a past scan already computed — its ratings summary, gate outcome, issue count, and full issue list.

1. Open the Quality tab (or the scan you're interested in).
2. Choose Analyze and reference the scan by its key or ID.
3. StretchDev returns the stored summary: the metrics, the gate result and whether it passed, the issue count, and the issues themselves.

Because it reads stored history, this view works even when the analysis engine is turned off — there's nothing to compute, only to retrieve.

Preview a quality gate

Use this to check whether a project's gate policy passes against its most recent scan's metrics, without running a fresh scan.

1. On the Quality tab, select the project.
2. Choose the gate preview.
3. StretchDev evaluates the project's policy against the latest scan's metrics and returns the metrics, the policy, and a Pass/Fail breakdown of each condition.

This is the safe way to test a policy change: adjust the gate, preview it against real recent metrics, and see the outcome before you commit. Previewing recomputes the gate, so it needs the engine enabled.

Score diff risk

Use this to score a proposed change described entirely by numbers — files changed, additions, deletions, complexity delta, hotspots touched, tests changed. It returns a risk score, churn, factors, and a recommended review path, and writes nothing. This is the same computation the Release-risk calculator uses; see Scoring release risk before you ship for the full walkthrough.

Analyze vs. run: which do I want?

• Run a scan when you want to record a new measurement of current code and keep it in history (it writes a scan and issues).
• Analyze / preview / score when you want an answer right now without adding to history — checking a past result, testing a gate change, or sizing up a change before it exists.

Example

You're considering tightening the Billing API gate to require Security A. Instead of guessing, you open the gate preview: it evaluates the new policy against the latest scan (Security C) and reports a failed condition. You now know the project would immediately fail, so you fix the underlying vulnerability first, re-scan, then apply the stricter gate — no throwaway scans cluttering your history.

Tips

• Reach for analyze a stored scan to revisit a result without re-running anything — it's the lightest, always-available view.
• Use the gate preview as your policy sandbox: change the gate, preview, repeat.
• Remember the split: analyze stored needs nothing; gate preview and diff-risk recompute and need the engine.
• None of these views change your data, so they're safe to use freely.

Troubleshooting

• "This project has no scans to evaluate yet." A gate preview needs at least one stored scan for its metrics — run a scan first.
• "Scan not found." The referenced scan key/ID isn't in your organization; confirm the workspace.
• The preview or diff-risk reports the engine is off. Those two recompute results; ask an administrator to enable the engine. Reading a stored scan does not need it.