Replace custom YAML handling with built-in import and export

[rhennigan]

Summary

• Replace the custom YAML parser and serializer with built-in YAML import/export support available in Wolfram Language 15.0+.
• Preserve missing/empty-file behavior and source labels in parse failures; simplify parse errors because the built-in importer does not expose line diagnostics.
• Use explicit UTF-8 byte-array conversion for string helpers to avoid ImportString/ExportString Unicode issues observed in 15.0.1.
• Update YAML and generated skill formatting expectations and add tests for anchors, aliases, Unicode, and mixed containers.

Validation

• 825 tests passed across YAML.wlt, InstallMCPServer.wlt, AgentSkills.wlt, and AgentSkillsBuild.wlt.
• CodeInspector reported no issues in changed source and test files; git diff --check passed.
• The additional DeployAgentToolsSkills.wlt run had four remaining failures involving Windows path separator expectations and an unavailable chmod executable.

[Copilot]

🟢 Approval recommended

The implementation matches the stated migration and includes comprehensive coverage for compatibility, errors, Unicode, and built-in YAML features.

0 open findings

What changed in this PR

Replaces custom YAML parsing/serialization with Wolfram Language 15.0 built-ins while preserving error handling and Unicode behavior.

Changes:

• Uses built-in YAML import/export with explicit UTF-8 conversion for strings.
• Simplifies YAML parse errors and removes custom parser code.
• Updates and expands YAML and generated-skill tests.

File	Description
Kernel/​YAML.wl	Implements built-in YAML handling.
Kernel/​Messages.wl	Simplifies YAML parse errors.
Kernel/​AgentSkills.wl	Updates parser documentation.
Tests/​YAML.wlt	Updates expectations and adds feature coverage.
Tests/​InstallMCPServer.wlt	Updates generated YAML expectations.
Tests/​DeployAgentToolsSkills.wlt	Updates skill frontmatter expectations.
Tests/​AgentSkills.wlt	Updates generated skill snapshots.

🧠 Review effort: Balanced

---

💡 Add a code-review agent skill for context-aware, tailored reviews. Learn more in the docs.