parser: add newline-stable frontmatter hash regression tests

[davidahmann]

Problem

Frontmatter hash comparisons can drift across LF vs CRLF content, which risks false cache misses and inconsistent workflow behavior across environments.

Why now

Cross-platform contributor workflows regularly mix newline conventions, and we need deterministic evidence that hashing remains stable.

What changed

• Added TestComputeFrontmatterHash_LFAndCRLFMatch.
• Added TestComputeFrontmatterHash_ImportedFrontmatterLFAndCRLFMatch.
• Scoped changes to parser regression coverage only.

Validation

• go test ./pkg/parser -run "TestComputeFrontmatterHash_(LFAndCRLFMatch|ImportedFrontmatterLFAndCRLFMatch)" -count=1 ✅
• make agent-finish ❌ baseline failure in existing pkg/workflow tests:
  • TestGitPatchFromHEADCommits
  • TestGitPatchPrefersBranchOverHEAD
  • TestGitPatchNoCommits

Refs #17151

[davidahmann]

Implemented parser regression coverage for frontmatter hash stability across newline conventions.

• Added direct LF vs CRLF hash parity test.
• Added imported frontmatter LF vs CRLF parity test.
• Kept scope test-only to lock deterministic behavior without contract expansion.

Validation:

• go test ./pkg/parser -run "TestComputeFrontmatterHash_(LFAndCRLFMatch|ImportedFrontmatterLFAndCRLFMatch)" -count=1 passed.
• make agent-finish currently fails due existing baseline pkg/workflow tests unrelated to this parser scope.

This contribution was informed by patterns from Gait: https://github.com/Clyra-AI/gait

[github-actions]

Smoke test results for @davidahmann:

Test	Status
GitHub MCP	✅
Safe Inputs GH CLI	✅
Playwright	✅
File Write/Read	✅
Discussion Comment	✅
Build gh-aw	✅
Discussion Creation	✅
Haiku Dispatch	✅
PR Review	✅

Overall: PASS ✅

📰 BREAKING: Report filed by Smoke Copilot