feat: docs reference generates a skill catalogue and a page-type reference under docs/reference/ from the shipped defaults, checked and written in the origin repo only (#189)

Files changed:
- CHANGES.md
- README.md
- VERSION
- docs/reference/README.md
- docs/reference/commands/README.md
- docs/reference/commands/types-instructions-and-docs.md
- docs/reference/page-types.md
- docs/reference/skills.md
- instructions/dev/doc-pull-through.md
- tools/CONTRACT.md
- tools/README.md
- tools/chemenu/commands/docs_verify.py
- tools/chemenu/reference_doc.py
- tools/chemenu/tests/test_reference_doc.py

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SnAJ7Z3CpVD3PRbN73QtU2

tools/README.md

@@ -89,7 +89,7 @@ tools/
     filelock.py            an exclusive lock on an open file, flock on POSIX and msvcrt on Windows - the only module that imports either
     prerequisites.py       prerequisites.txt read from Python, plus the platform and long-path questions `doctor` asks
     install_doc.py         INSTALL.md held to the instructions: its prerequisites lists generated from prerequisites.txt, its setup questions matched to setup-instance.md's markers
-    reference_doc.py       the human command reference under docs/reference/: one page per command group, an overview and the exit-42 causes, rendered from the records without what addresses an agent
+    reference_doc.py       the human reference under docs/reference/: one page per command group, an overview and the exit-42 causes, rendered from the records without what addresses an agent; and, in the origin repo only, the skill catalogue and the page types, rendered from the shipped SKILL.md files and type-specs
     corpus_cache.py        one parsed corpus per commit, never cached while the tree is dirty
     kb_scan.py             page iteration/loading over kb/
     blocks.py              generated regions in a page body, found by marker rather than by heading