feat: consolidate reviewed Diagram Design 2.6.0 release (#134)

* fix(commands): stop hardcoding the visual-type count

Integrates the final reviewed diff from PR #112.

* docs(high-level): correct effective width checklist

Integrates the final reviewed diff from PR #105.

* fix(commands): quote profile argument hint

Integrates the final reviewed diff from PR #126.

* test(sankey): gate conservation and ribbon geometry

Integrates the final reviewed diff from PR #103.

* feat(tooling): add doctor diagnostics gate

Integrates the final reviewed diff from PR #132.

* feat(scripts): lint diagrams as rendered

Integrates the final reviewed diff from PR #18.

* feat(types): add quantitative polar chart

Integrates the final reviewed diff from PR #101.

* chore(release): consolidate reviewed changes in 2.6.0

Documents every combined gate and synchronizes Claude, Codex, and Factory metadata for the 39-type release.

* fix(doctor): support installed-skill projects

---------

Co-authored-by: mysoniw <soniw1@gmail.com>
Co-authored-by: thang <76550342+thangldw@users.noreply.github.com>
Co-authored-by: YinShaokun <yinshaokun@didiglobal.com>
Co-authored-by: 0xDarkMatter <0xDarkMatter@users.noreply.github.com>
Co-authored-by: Rupam Pal <rupam.pal@euromonitor.com>
Co-authored-by: Yuvraj3905 <yuvraj202001@gmail.com>

docs/adr/0002-semantic-patterns-do-not-expand-the-taxonomy.md

@@ -1,6 +1,6 @@
 # ADR 0002 — Semantic patterns never expand the visual-type taxonomy
 
-**Status:** accepted (v2.3)
+**Status:** accepted (v2.3; amended v2.6)
 
 ## Context
 
@@ -8,7 +8,7 @@ Auditing behavior-rich figures (queues, policy traces, trust boundaries) showed
 
 ## Decision
 
-Behavior is a separate axis. The seven semantic patterns in `references/semantic-patterns.md` each route to the **nearest existing visual type** for layout; a pattern owns semantic primitives and a tighter budget, never a second layout grammar. The visual-type count only moves when a genuinely new *layout* grammar appears (it did in v2.5.10 — see [ADR 0007](0007-new-layout-grammars.md), 28 → 38).
+Behavior is a separate axis. The seven semantic patterns in `references/semantic-patterns.md` each route to the **nearest existing visual type** for layout; a pattern owns semantic primitives and a tighter budget, never a second layout grammar. The visual-type count only moves when a genuinely new *layout* grammar appears (it did in v2.5.10 — see [ADR 0007](0007-new-layout-grammars.md), 28 → 38 — and again for Polar, 38 → 39).
 
 ## Consequences
 
@@ -20,6 +20,8 @@ Behavior is a separate axis. The seven semantic patterns in `references/semantic
 
 **2026-08-18 — the count is 28.** Treemap was admitted under the escape clause above: recursive area subdivision is a layout grammar no existing type provides (bar encodes with length, nested with containment and no quantity, pyramid with rank). It shipped the full §10 set, and the counters named above moved 27 → 28 together with the prose.
 
-The decision itself is unchanged — semantic patterns still never add a type, and the count still moves only for a new *layout* grammar. What this amendment records is the procedure: the two counters are this ADR's enforcement, so a PR that edits them without amending this file has quietly made itself the authority. Amend here in the same PR, or the number in the test is just whatever the last contributor typed.
-
 **2026-08-19 — the count is 38.** Ten additional visual grammars were admitted under the same escape clause: Sankey, fishbone, Wardley map, kanban, user journey, deployment, dependency graph, UML class, story map, and database schema. The per-type argument is in [ADR 0007](0007-new-layout-grammars.md); each ships the full §10 set, and the two counters move 28 → 38 together with the prose.
+
+**2026-08-20 — the count is 39.** Polar was admitted under the same escape clause: angle encodes ordered cyclic categories and linear radius encodes one quantitative series, a layout grammar no existing type provides. It shipped the full §10 set, and the counters named above moved 38 → 39 together with the prose.
+
+The decision itself is unchanged — semantic patterns still never add a type, and the count still moves only for a new *layout* grammar. What this amendment records is the procedure: the two counters are this ADR's enforcement, so a PR that edits them without amending this file has quietly made itself the authority. Amend here in the same PR, or the number in the test is just whatever the last contributor typed.

docs/adr/0007-new-layout-grammars.md

@@ -61,4 +61,8 @@ Kept out on purpose, so the next audit does not relitigate them.
 - **ER's scope was narrowed, not duplicated.** `type-er.md` now states that it is entity-level and points at `type-db-schema.md` for the physical schema. Without that edit the two types would have overlapping "Best for" claims and the router would be ambiguous.
 - The ADR 0004 byte cap became binding. Adding these types required trimming SKILL.md body prose: the §11 import consequences were compressed, the terminal-variant and typography paragraphs were tightened, and §4's six connector anti-pattern rows collapsed into one row pointing at §6 — that table was the third statement of rules already given in full in §6 and checked one by one in §9. `SKILL.md` sits at ~39.2 KB against the 40 KB cap. The next type must be paid for the same way, never by trimming the frontmatter description.
 - Per-type budget rows stayed in SKILL.md §7 rather than moving into the type references, because several older references (venn, pyramid, layers, ER, swimlane, timeline) do not state their own limits. Moving the rows out would have lost those numbers.
-- `docs/screenshots/` has no PNGs for the new types yet; the README lists them in a text table instead of the image grid until the screenshots are generated.
+- At acceptance time, `docs/screenshots/` had no PNGs for the new types; README listed them in a text table pending generation.
+
+## Amendment — canonical screenshots shipped
+
+The ten types now have canonical PNGs, image-grid entries in README, and source/screenshot digests in `docs/screenshots/manifest.json`. Polar was admitted separately afterward and moved the repository-wide count from 38 to 39; this ADR still records the ten-type 28 → 38 decision.