Passing checks still need meaningful coverage.
202 illustrative checks pass; payment and AI recovery remain outside the enforced test set.
Read illustrative evidence →Asterpath · Revision 4
Asterpath / sample evidence pack
A sample audit you can explore.
Read the assessment, inspect selected supporting evidence and carry the priorities into another engineer’s workflow. Everything stays together in this offline pack.
Fictional, not a scanner result. FICTIONAL ILLUSTRATIVE SAMPLE. Asterpath, its product, evidence, findings and measurements are invented. No Asterpath application was inspected and no scanner was executed. These are original GadirLabs illustrations, not native vendor exports.
Start with the decision
01 / THE OWNER’S VIEW
Understand what matters first.
The sole launch blocker is paid access after an interrupted update. Read the decision, then the action plan and retest outcomes.
Executive decision · page 302 / THE ENGINEER’S VIEW
Trace the reasoning to the evidence.
Use stable finding and evidence IDs to connect the report, structured examples and editable priorities.
Priorities and ownership · page 64See what sits behind the interface.
Authentication, backend behaviour and database quality have their own chapters. The coverage guide also makes email delivery, business rules, capacity, safe support, continuity and AI quality easy to locate, with evidence limits stated explicitly.
Explore assessment coverage →SELECTED SUPPORTING DETAIL
Explore the evidence behind the assessment.
Each readable page has a matching JSON illustration. These are two representations of one record, not extra findings.
Triage unused-code candidates before calling them defects.
Twelve fictional candidates become four cleanup items after entry-point and generated-code review.
Read illustrative evidence →Explain the mobile wait with a bounded lab example.
The fictional landing page has a mobile LCP of 5.1 seconds and a desktop LCP of 2.0 seconds.
Read illustrative evidence →Follow the customer through interruption and recovery.
Normal journeys work in the fiction; interrupted payment, AI timeout and dialog-close paths reveal different gaps.
Read illustrative evidence →Use manual evidence to explain the access barrier.
The fictional export flow remains possible, but loses the keyboard user's position when the dialog closes.
Read illustrative evidence →Reconcile provider cost with successful customer work.
The same invented billing period contains £351 in the product ledger and £428 in the provider invoice.
Read illustrative evidence →Give the first task a clear visual priority.
Four equally prominent workspace actions compete for attention in the illustrated design.
Read illustrative evidence →Distinguish sending an alert from delivering it.
The fictional relay reports local success even when the destination rejects the notification.
Read illustrative evidence →Take the detail into your own workflow.
Editable findings
All eight findings, with severity, priority, owner, dependencies and retest outcomes.
Scan and evidence register
Explicit actual status: Not run. Shows what each fictional record represents.
File inventory and provenance
SHA-256 hashes, sizes, record kinds and evidence IDs for the packaged payload.
Reading and verification notes
How to use the files, check integrity and understand the sample boundaries.
Where scanner exports would fit.
The following pages explain two optional tools and their export limitations. They contain no invented vendor results.
SonarQube
First-party code quality, security, edition limits and permitted processing of output.
Read SonarQube availability note →JFrog Xray
Dependency risk, advanced scanning options and genuine native exports.
Read JFrog Xray availability note →What a real client handover adds
Actual candidate-linked evidence, permitted native exports from completed checks, manual observations and triage. Unavailable or withheld evidence has a stated reason. This public sample contains no private product code, real invoices, internal audit checklist or vendor artwork.