chore(catalog): bump v-model extension to v0.7.1

[leocamello]

Updates the v-model entry in extensions/catalog.community.json to point at the new v0.7.1 release.

v0.7.0 was never submitted upstream, so this single bump catches the catalog up from 0.6.0 directly to 0.7.1. Follows the same pattern as the previous bumps:

• v0.2.0 → Update V-Model Extension Pack to v0.2.0 #1656
• v0.3.0 → Update V-Model Extension Pack to v0.3.0 #1661
• v0.4.0 → Update V-Model Extension Pack to v0.4.0 #1665
• v0.5.0 → Update V-Model Extension Pack to v0.5.0 #2108
• v0.6.0 → chore(catalog): bump v-model extension to v0.6.0 #2399

Catalog changes

Field	Before	After
version	0.6.0	0.7.1
download_url	…/v0.6.0.zip	…/v0.7.1.zip
provides.commands	14	17 (+3 bridge commands)
provides.hooks	1	4 (+3 lifecycle hooks)
stars	21	27 (live count from GitHub API)
updated_at	2026-04-25	2026-05-17 (v0.7.1 release date)

Top-level catalog updated_at bumped 2026-05-15 → 2026-05-17.

What's new since v0.6.0

v0.7.0 — Bridge Commands (catalog-relevant)

Three new commands and three new lifecycle hooks connect V-Model specification artifacts to executable implementation, completing the spec-to-code gap:

Command	Wraps	Adds
/speckit.v-model.plan	/speckit.plan	V-Model-enriched plan.md synthesis with additive HTML-comment traceability; pinned-schema validation; structured run summary.
/speckit.v-model.tasks	/speckit.tasks	TDD-ordered task decomposition consuming unit-test.md / integration-test.md / system-test.md / acceptance-plan.md; hazard-driven priority elevation; per-task Implements-directive headers.
/speckit.v-model.implement	/speckit.implement	Pre-flight V-Model gate (run-v-model-gate.sh); four-level test generation; sentinel-managed splicing into existing sources; Implements-directive hallucination guard; post-implement trace hook.

New hooks: after_specify → v-model.requirements; before_implement → v-model.trace; after_implement → v-model.trace. Plus eight validator/MF hardenings (status gate, domain profile, schema validator, splicer hardening, etc.).

v0.7.1 — Bridge Command Reconciliation & Governance

Reconciles the v0.7.0 bridge commands with their user-facing intent — that /speckit.v-model.implement produces a fully working and validated implementation from the V-Model artefact set:

• All 8 V-Model artefacts mandatory for implement. No graceful skip on missing test plans or design artefacts; iterates module-design.md directly per the V-Model design; explicit test-first emphasis pulling all V-Model artefacts into code-gen context.
• /speckit.v-model.plan and tasks clarified as OPTIONAL. The Direct Path via /speckit.v-model.implement is self-sufficient; the optional bridges exist for users mixing V-Model with spec-kit-core's downstream toolchain.
• Real contributor operating manual. New AGENTS.md (replaces auto-generated CLAUDE.md); constitution v1.0.0 → v1.1.0 with additive Bridge Command Discipline, V-Model Artifact Map, and expanded ID schema.
• Canonical-ID grammar generalised. Accepts any 2–3-letter uppercase sub-prefix on any root (REQ-DR, REQ-LC, REQ-SEC, SYS-DR, etc.), matching the project's actual conventions.

No new commands or hooks in v0.7.1.

Verification

Suite	Result
python3 -c "import json; json.load(open('extensions/catalog.community.json'))"	✅ valid JSON
Spec-kit-v-model pytest structural	122 passed, 0 failed
Spec-kit-v-model BATS	457 passed, 0 failed
Spec-kit-v-model LLM-judge	10/10 effectively passing (1 LLM-variance flake passes on re-run)
Download URL HEAD	HTTP 302 (matches v0.7.0's verified-redirect pattern)

Full release narratives:

• v0.7.0 CHANGELOG
• v0.7.1 CHANGELOG