Fix documentation drift in README and CONTRIBUTING

[dmccoystephenson]

Summary

A documentation accuracy sweep was performed against the current source of truth (LICENSE, build.gradle/settings.gradle output, git remote), and three drift points were found and corrected:

• The git clone URL in README's Installation section pointed at Preponderous/Parpt, an org that does not match the actual remote (Stephenson-Software/Parpt) — a new contributor following the steps literally would clone the wrong (or a nonexistent) repository.
• The java -jar commands in README's Installation and Getting Started sections referenced parpt.jar / build/libs/parpt.jar, but the Spring Boot build actually produces build/libs/Parpt-<version>.jar (verified by running ./gradlew bootJar and inspecting build/libs/). Both references were updated to the glob form build/libs/Parpt-*.jar so the doc does not go stale again on the next version bump.
• CONTRIBUTING's License section stated contributions are licensed under "the MIT License," which conflicts with the repository's actual LICENSE file and the license section of README (Stephenson Software Non-Commercial License). The line was corrected to reference Stephenson-NC and link to LICENSE.
• CONTRIBUTING's CI description cited "currently 61 tests," which no longer matches the actual count (58, confirmed via grep -rn "@Test" src/test | wc -l); the figure was updated.

No source/behavior changes are included — this PR is docs-only.

This PR description was drafted during a Gardener session (https://github.com/Stephenson-Software/gardener).

Test plan

• ./gradlew test run locally — all 58 tests pass (baseline confirmed unaffected, since only Markdown files were touched)
• ./gradlew bootJar run locally to confirm the actual jar filename referenced in the corrected README matches real build output
• Every corrected claim (org name, jar filename, license, test count) checked against source (git remote, build/libs/ contents, LICENSE file, test grep) rather than assumed

[dmccoystephenson]

Self-review rubric (this PR is a Stage A documentation accuracy sweep — no source code touched):

• Scope: PASS — every changed line is a documentation string; each of the two files touched is directly implicated by one of the four drift points listed in the PR body. No unrelated formatting or renames.
• Tests-new: N/A — no new public methods/functions introduced (docs-only change).
• Tests-fix: N/A — not a bug fix; regression-test-via-revert does not apply to prose corrections. Each claim was instead verified empirically against source before editing (see below), which is the applicable evidence standard for a docs sweep.
• Sibling structure: PASS — no new files created; existing README/CONTRIBUTING section structure and heading hierarchy preserved as-is.
• Sibling renames: N/A — no identifier renames in this PR.
• Docs: PASS — this PR is the doc-accuracy fix; each item below shows the verification performed:
  • Clone URL — confirmed actual remote is Stephenson-Software/Parpt via git remote -v; README previously said Preponderous/Parpt.
  • Jar filename — ran gradlew bootJar and inspected build/libs/; actual output is Parpt-0.2.0-SNAPSHOT.jar (derived from settings.gradle's rootProject.name plus build.gradle's version), not parpt.jar. Used the glob form Parpt-*.jar in both README commands so the doc doesn't re-drift on the next version bump.
  • License mismatch — compared CONTRIBUTING.md's "MIT License" claim against the actual LICENSE file and README's own License section, both of which state Stephenson-NC; CONTRIBUTING corrected to match and linked to LICENSE.
  • Test count — counted @test annotations under src/test recursively: 58, vs. CONTRIBUTING's stale "61".
• Issue resolution: N/A — no tracking issue exists for this drift; found during triage of the repo's docs against source. Per Stage A, an issue is only required when the code — not the docs — is what's wrong; here the docs were wrong, so fixing them directly is the correct path.
• manual validation: PASS — gradlew test run on the PR head, 58/58 tests pass (BUILD SUCCESSFUL). Expected to be a no-op signal here since no source file changed, but confirmed rather than assumed.

Repo-specific rubric items (shell command tests, round-trip coverage, I/O behind an interface): N/A — no command, domain, or persistence code was touched.

Summary: docs-only, tightly scoped (2 files, 5 lines changed each direction), every corrected claim checked against source rather than memory. No judgment-call items to flag for follow-up review.

This comment was drafted during a Gardener session (https://github.com/Stephenson-Software/gardener).