docs(skill): fix 5 factual mistakes flagged in self-review

All five were verified against the real openkb code and live CLI
output before fixing. Net effect: an agent following the skill no
longer hallucinates commands or off-by-one page reads, and the
reference docs match what `openkb list` / `index.md` / summary
frontmatter actually contain.

1. SKILL.md no longer references a nonexistent `openkb where` —
   uses `openkb status` everywhere (the rest of the skill already
   did; line 67 was a stale fragment from an earlier iteration).

2. SKILL.md's `jq` long-doc-page instruction is now unambiguous:
   `jq '.[N-1]' wiki/sources/<doc>.json` (N = 1-indexed PDF page
   number). Previous "(page N, 0-indexed)" wording risked
   systematic off-by-one in agent reads.

3. references/commands.md shows the real `openkb list` Type column:
   `pageindex` (for `long_pdf` registry entries) and `short` (for
   every other format), matching `_TYPE_DISPLAY_MAP` in cli.py. The
   prior "long_pdf / md" example was the raw registry values, not
   the displayed ones.

4. references/wiki-schema.md's index.md example now shows `(short)`
   / `(pageindex)` type tags (what the compile pipeline emits), not
   the raw `(long_pdf)` / `(md)` placeholders.

5. references/wiki-schema.md's summary frontmatter `full_text:` is
   no longer described as "short docs only" — long PDFs also have
   the field, pointing at the `.json` paginated content.

The remaining `long_pdf` strings in the docs intentionally describe
the raw registry value in `.openkb/hashes.json`, contrasted with the
displayed `pageindex` — they're the correct framing now.

Author: KylinMountain

skills/openkb/SKILL.md

@@ -64,15 +64,15 @@ After capturing the KB path from `openkb status`, drill in via:
 
 ## Read content
 
-(Paths shown relative to the KB root from `openkb where`. Prepend it
-in real calls.)
+(Paths shown relative to the KB root captured from `openkb status`'s
+first line. Prepend it in real calls.)
 
 | Goal | How |
 |---|---|
 | Read a concept page | `Read <kb>/wiki/concepts/<slug>.md` |
 | Read a document's summary | `Read <kb>/wiki/summaries/<doc>.md` |
 | Read a short doc's full text | `Read <kb>/wiki/sources/<doc>.md` |
-| Read a long doc's specific page | `jq '.[N]' <kb>/wiki/sources/<doc>.json` (page N, 0-indexed) |
+| Read a long doc's specific page | `jq '.[N-1]' <kb>/wiki/sources/<doc>.json` (where N is the 1-indexed PDF page number; `.[0]` is page 1) |
 | Get a synthesized answer across sources | `openkb query "<question>"` |
 | Find an exact phrase | `Grep -r "<phrase>" <kb>/wiki/` |
 | Follow a `[[wikilink]]` | `Read` the linked path under `<kb>/wiki/` |

skills/openkb/references/commands.md

@@ -13,7 +13,7 @@ $ openkb list
 Documents (2):
   Name                                     Type         Pages
   ---------------------------------------- ------------ --------
-  paper.pdf                                long_pdf     42
+  paper.pdf                                pageindex    42
   notes.md                                 short
 
 Summaries (2):
@@ -28,9 +28,15 @@ Concepts (5):
   - multi-head-attention
 ```
 
-- `Type` shows the registry's `type` field: `long_pdf` for
-  PageIndex-indexed PDFs, otherwise the file extension (`md`,
-  `docx`, `pdf`, …).
+- `Type` is the *display* form of the registry's `type` field, mapped
+  through `_TYPE_DISPLAY_MAP`:
+    - PageIndex-indexed long PDFs (registry `type: long_pdf`) display
+      as `pageindex`.
+    - Every other format (`md`, `docx`, `pdf` short, `txt`, …) displays
+      as `short`.
+  The raw registry value lives in `.openkb/hashes.json`; the displayed
+  value is what surfaces in `openkb list` and in `index.md` type tags
+  (`(short)` / `(pageindex)`).
 - `Pages` only populated for long PDFs.
 - The Summaries and Concepts lists are simply directory listings of
   `wiki/summaries/` and `wiki/concepts/` minus their `.md` suffix.

skills/openkb/references/wiki-schema.md

@@ -46,8 +46,8 @@ Plain Markdown with three top-level sections:
 # Knowledge Base Index
 
 ## Documents
-- [[summaries/paper]] (long_pdf) — Brief from the summary frontmatter.
-- [[summaries/notes]] (md) — ...
+- [[summaries/paper]] (pageindex) — Brief from the summary frontmatter.
+- [[summaries/notes]] (short) — ...
 
 ## Concepts
 - [[concepts/attention]] — Brief from the concept frontmatter.
@@ -57,6 +57,11 @@ Plain Markdown with three top-level sections:
 - [[explorations/some-saved-query]] — User's saved query answer.
 ```
 
+The type tag in parentheses is always either `(short)` or
+`(pageindex)` — never the file extension. Short = anything the
+markitdown path can convert (md, docx, html, txt, short PDFs);
+pageindex = a long PDF indexed by PageIndex.
+
 Section headings are kept even when empty (e.g. after removing all
 documents the `## Documents` heading stays). Entry order is roughly
 insertion order, not alphabetical.
@@ -70,10 +75,15 @@ Per-document summary. Frontmatter:
 sources: [raw/paper.pdf]        # The original ingested file
 brief: One-line description.
 doc_type: short                  # short | pageindex
-full_text: sources/paper.md      # short docs only — link to the source
+full_text: sources/paper.md      # short docs: .md ; long PDFs: .json
 ---
 ```
 
+`full_text` always points at the converted source file: short docs
+get `sources/<name>.md` (markitdown output); long PDFs get
+`sources/<name>.json` (per-page content array — see the long-doc
+section below for how to read it).
+
 Body is the LLM-synthesized summary plus a `## Related Concepts`
 section linking to the concepts this doc touches.