Skip to content

chore: retain Rust/Tauri plugin integration experiment - #83

Merged
jdylanmc merged 2 commits into
mainfrom
jdylanmc/rust-plugin-spike
Oct 2, 2026
Merged

jdylanmc merged 2 commits into
mainfrom
jdylanmc/rust-plugin-spike

Conversation

@jdylanmc

@jdylanmc jdylanmc commented Sep 30, 2026 •

Copy link
Copy Markdown
Owner

Purpose

Retain the Rust/Tauri integration spike at a useful, reproducible stopping
point. This is an experimental tool and findings package, not production
Rust support or a public plugin SDK.

Refs #1. The real-repository adoption work informs #46; it does not close the
human acceptance requirement.

Included

  • Isolated experiments/plugin-spike: explicit language/framework contribution
    contract, existing TypeScript adapter, syntax-backed Rust analysis, and
    bounded Tauri command bindings.
  • Source snapshots, per-capability limitations, explicit unresolved references,
    checked source-derived/source-traced explanations and a separate topo-spike
    CLI with project guidance.
  • Archify and Graphviz rendering experiments with accurate engine-specific
    receipts; no renderer substitution hidden as a successful Archify check.
  • Portable setup, 80 focused tests, actual packed-consumer smoke and a dedicated
    Node 22 macOS/Linux workflow.
  • Concise integration findings and measured evidence, including a small
    synthetic semantic-probe fixture. Full local generated artifacts, packages,
    private paths and caches are excluded.

What we learned

The shared graph/evidence model is usable, but scanner capabilities, source
anchors, framework joins and coverage need deliberate language-neutral
contracts. Rust syntax extraction is useful without builds; it does not resolve
all calls, conditional compilation, macros, or runtime behavior.

The controlled rust-analyzer probe demonstrated selected definitions,
references and call hierarchy under an explicitly limited project/environment.
It is historical feasibility evidence, not a shipped semantic backend or
an operating-system sandbox.

Fresh-agent use exposed real source-binding, CLI-entrypoint, publication and
diagram-layout defects. The retained implementation includes the fixes and
regressions, while preserving the limits and failed approaches in the findings.

Deliberately not included

  • No changes to production packages/, the existing production CI or npm
    publication workflow.
  • No new production Rust/Tauri capability, plugin marketplace, unrestricted
    build/procedural-macro execution, or universal determinism claim.
  • No modification to PR-Sniper application source.
  • No branded core initialization landing/catalogue fix. The prototype book
    starts at its first diagram; this is not the requested Storybook-like
    application shell. That foundational fix is the next separate delivery.

Evidence and validation

  • Local Node 22: 86/86 experiment tests passed, including book-publication,
    reserved-ID and nonempty-verification regressions.
  • Real packed-consumer installation, npm-bin execution, preserved init context,
    partial scan (exit 2), source references, and both renderer paths passed.
  • Historical PR-Sniper evidence: source-change rejection/repair/restoration,
    144 original context files preserved, five substantive technical views,
    336 verified excerpts across 19 files, and bounded browser measurements.
  • Normal public-registry install, experiment tests and packed-consumer proof:
    Node 22 macOS and Linux passed.
  • Existing production regression and installed-toolbelt workflow:
    passed.
  • Independent whole-deliverable review completed; all three curation findings
    corrected and independently reverified at
    6c0dc0ea4cdd00e90e14b61d862effcb5ab6fc6e.

See README,
REPORT, and
evidence/summary.json.

The reviewed head contains the current observed main baseline
b64bb2fc9c1ba6319287f7c701588eee3de2d1d5.
Human review remains outstanding. No merge or npm publication is authorized.

@jdylanmc
jdylanmc marked this pull request as ready for review September 30, 2026 16:07
@jdylanmc
jdylanmc merged commit 321a405 into main Oct 2, 2026
3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant