feat(model-overlays): add sonnet-5 overlay + resolver mapping

[chrisquorum]

What

Claude Sonnet 5 (claude-sonnet-5) is the current Sonnet-tier flagship — near-Opus
quality on coding and agentic work. Today claude-sonnet-5* resolves through
models.ts to the generic claude overlay. This adds a dedicated sonnet-5
overlay so it inherits the claude base plus Sonnet-5 family nudges.

Changes

• model-overlays/sonnet-5.md (new) — {{INHERIT:claude}} + three nudges: literal
  instruction following (state scope explicitly), scope work to the request (raise
  effort rather than prompting around shallow reasoning at low effort), and
  verbosity that tracks task complexity.
• scripts/models.ts — add sonnet-5 to ALL_MODEL_NAMES and a resolveModel
  mapping (/^claude-sonnet-5(-|$)/ → sonnet-5), placed before the generic claude
  fallback. claude-sonnet-4-6 still falls through to generic claude.
• test/model-overlay-sonnet-5.test.ts (new) — gate-tier test mirroring the
  opus-4-* overlay tests.

Notes

• Nudges are drawn from documented Sonnet 5 behavioral shifts (literal instruction
  following, effort-scoped work, complexity-tracked verbosity), kept to the short
  form the other overlays use. Retune as more patterns emerge.

Validation

• resolveModel('claude-sonnet-5') / 'claude-sonnet-5-sol' → sonnet-5;
  claude-sonnet-4-6 → claude, claude-opus-4-7, and bare claude unchanged.
• bun test test/model-overlay-sonnet-5.test.ts → all pass.
• bun run gen:skill-docs → no generated-doc churn (only the three files above).
• bun run typecheck → no new errors in the changed files.

[trunk-io]

Merging to main in this repository is managed by Trunk.

• To merge this pull request, check the box to the left or comment /trunk merge below.

After your PR is submitted to the merge queue, this comment will be automatically updated with its status. If the PR fails, failure details will also be posted here