At a glance
| Path | Visible action | Captured evidence | Important boundary |
|---|---|---|---|
| Terminology baseline | Create baseline | Term identity, name, status, version, space, and scanned page state | It stores page versions, not page content hashes. Trace-link coverage is unavailable. |
| Terminology baseline | Baseline evidence report | Index coverage, usage coverage, change counts, and data-quality warnings | Term state includes every status. The approved count is separate, and capture stops at 10,000 terms. |
| Terminology baseline | Changed since baseline | Changed terms, changed pages, deleted terms, and pages with new findings | This comparison uses current state and recorded page deltas. |
| Enterprise freeze | Prepare and queue freeze | Approved meanings, exact matching page content, revisions, SHA-256 hashes, usage, findings, and retention. | Selected scan rows must share one glossary snapshot. Capacity admission runs before queueing. |
| Enterprise freeze | Frozen meanings and revisions | Availability, conformance, impact, coverage blockers, and retained trace links when present | Missing evidence remains visible. Detailed verification data is not rendered as a report. |
| Local qualification | No production interface | A deterministic workload of 250,000 mixed artifacts | It is local simulation evidence, not Forge staging or production capacity evidence. |
How it works
- Open Terminology Baseline.
- Use Create baseline for a versioned comparison point across selected spaces.
- Review Baseline evidence report and Changed since baseline.
- Use Prepare and queue freeze when exact content, hashes, retention, or richer evidence is required.
- Review Freeze progress, Coverage blockers, and Frozen meanings and revisions.
- Select Run verification before exporting the enterprise package.
Example
At a release checkpoint, an administrator creates a terminology baseline for two governed spaces. It records term versions and scanned page versions. If exact page content and hashes are required, the administrator creates an enterprise freeze. Later edits appear through comparison or new evidence. They do not alter the captured point.
What this does not mean
Neither path proves that source content is correct or compliant. The legacy 200,000-page code cap is an input ceiling, not production capacity evidence. The 250,000-artifact qualification is a deterministic local simulation. It does not establish Forge latency, consumption, cost, residency, or production capacity. Application hashes are tamper evidence, not legal WORM storage.
Related questions
Next step
Choose the evidence depth first. Use Create baseline for version comparison. Use Prepare and queue freeze for exact content, hashes, retention, or trace evidence. View Compliance Glossary on Atlassian Marketplace.
View on Atlassian Marketplace