test: compare a fresh full baseline with PR #5 incremental analysis

[ivanmilevtues]

Full-analysis comparison with #5

A fresh full baseline has been generated and committed. This draft PR targets refactor/vet-directory-review-demo, not main, and contains the same application source as #5.

• Full run (passed in 3m5s): https://github.com/CodeBoarding/spring-petclinic/actions/runs/34265528374
• Full graph: .codeboarding/analysis.json in this PR
• Download full baseline: https://github.com/CodeBoarding/spring-petclinic/actions/runs/34265528374/artifacts/10071755610
• Original incremental review: refactor: extract veterinarian directory queries for review demo #5 (comment)
• Download original incremental graph: https://github.com/CodeBoarding/spring-petclinic/actions/runs/34256446154/artifacts/10068195857

Controls and verification

The full-analysis workflow pins the exact action revision from the original review, including engine 0.14.0, and uses the same hosted provider and depth 2. It runs mode: sync with force_full: true on this comparison branch only. The moving v1 tag has changed since the original run, so it is deliberately not used here.

Logs confirm Running FULL analysis workflow and Force full analysis: skipping static analysis cache. A source diff against the original PR branch is empty. The original artifact metadata records mode: incremental and seed_source: base.

Both outputs have the same five top-level component names and eight component relations; that is a coarse comparison, not a claim that the complete graphs are identical. Hosted model responses can vary between runs.

Left as a draft to avoid an unrelated incremental review of generated baseline files. No application code changes, no changes to main, and no modifications to the original review result. This PR is for comparison, not merging.

[ivanmilevtues]

Full-analysis Mermaid — comparison with PR #5

Generated directly from the committed analysis.json. This is the full snapshot, so no added/modified/deleted colors are applied.

Top level — same depth as the PR #5 comment

graph LR
    n_1["owner"]
    n_2["vet"]
    n_3["system"]
    n_4["model"]
    n_5["Loose files in org.springframework.samples.petclinic"]
    n_1 -- "calls" --> n_4

Loading

Expanded — all stored components and relationships

graph LR
    subgraph n_1["owner"]
        n_1_1["Owner"]
        n_1_2["PetController"]
        n_1_3["Pet"]
        n_1_4["Visit"]
    end
    n_2["vet"]
    n_3["system"]
    n_4["model"]
    n_5["Loose files in org.springframework.samples.petclinic"]
    n_1_1 -- "calls" --> n_1_3
    n_1_1 -- "calls" --> n_4
    n_1_2 -- "calls" --> n_1_1
    n_1_2 -- "calls" --> n_1_3
    n_1_2 -- "calls" --> n_4
    n_1_3 -- "calls" --> n_4
    n_1_4 -- "calls" --> n_1_1
    n_1_4 -- "calls" --> n_1_3

Loading

The eight stored relationships collapse to one cross-component edge (owner → model) at the top level; the others are internal to owner. The JSON contains no inter-component relationships for vet, system, or the loose-files component, so none are invented here.

Compare with PR #5’s incremental diagram. The top-level names and edge match; PR #5 additionally marks vet as modified.

Full-analysis run · same application source, action revision, engine 0.14.0, hosted provider, and depth 2 as the original run.