@event4u/agent-config 14.18.0 → 14.19.0

package/dist/scripts/_lib/model_tier.js

@@ -11,7 +11,7 @@
  * - the repo generator (`condense.py generate_claude_skills`), which
  *   builds the package's own `.claude/skills/` tree, and
  * - the consumer install finalizer
- *   (`install.py finalize_claude_model_tiers`), which rewrites the
+ *   (the installer's `finalize_claude_model_tiers`), which rewrites the
  *   installed `.claude/skills/` tree so Claude Code performs the
  *   per-turn model switch on a consumer with `model.auto_switch: auto`.
  *

package/docs/guidelines/agent-infra/evaluator-independence-mechanics.md

@@ -51,7 +51,7 @@ drew.** It said `pre_tool_use` "exists on three hosts" and that the guard has
 tree. **Downward:** the manifest's own `native_event_aliases` table already maps
 `preToolUse` (cursor), `PreToolUse` (cline) and `BeforeTool` (gemini) onto
 `pre_tool_use`, so there the guard is **unbound, not unbindable**; only windsurf
-and copilot carry no pre-tool surface at all. **Upward, and worse:**
+and copilot carry no pre-tool alias row at all. **Upward, and worse:**
 `host_semantics.ts` certifies **claude alone**, and the augment and cowork
 trampolines discard dispatcher output and `exit 0` unconditionally — so on two
 of the three "enforcing" hosts this guard runs and is then ignored. Nor may the

package/docs/guidelines/agent-infra/roadmap-progress-mechanics.md

@@ -19,7 +19,7 @@ _Origin: migrated from `.agent-src.uncondensed/rules/roadmap-progress-sync.md` p
 >    stale dashboard even if local hooks were bypassed.
 >
 > Hook is primary; the prose below is the specification the hook
-> implements and the fallback when the platform has no hook surface.
+> implements and the fallback where no hook is bound on the platform.
 
 ## Iron Law — dashboard sync
 

package/docs/guidelines/design-antipatterns.md

@@ -68,105 +68,128 @@ generation ships a visibly different default aesthetic. It is a *when-does-it-
 rotate* note, not a new pattern — never add a tell here that lacks a catalog
 entry in the sections below.
 
+### The three census columns — class, remediation, verification
+
+`Class` says what KIND of rule an entry is; **status** (§ Detector status) says
+how its enforcement is wired. The two axes are independent, and a reader who
+conflates them draws the wrong conclusion in both directions: a `floor` may be
+`judgment-only` because nothing can detect it yet, and a `style-preference` may
+be `backed` because it happens to be trivially greppable. Detectability is not
+authority.
+
+- **Class** — one of `floor` (an objective threshold, accessibility or
+  correctness) · `invariant` (a structural rule that must hold) ·
+  `craft-presumption` (a default a competent designer usually follows,
+  overridable on evidence) · `style-preference` (taste) ·
+  `reference-constraint` (bound to a supplied reference, brand, or `DESIGN.md`).
+  The two soft classes lose to a supplied reference artifact, a coherent
+  incumbent, and a project `DESIGN.md` — asserted on the resolver in
+  `src/scripts/_lib/ui_authority.ts` rather than stated here as prose.
+- **Remediation** — the order to try, worst-first: `delete` before `reduce`
+  before `retune`, with `replace` where the pattern has a direct substitute.
+- **Verification** — how a fix is checked: `static` (readable from the source)
+  · `render` (needs the rendered surface) · `feel` (perceptual, per the `feel`
+  evidence type in `docs/contracts/evidence-artifact-types.md`) · `judgment`.
+
 ### Consistency Locks & layout caps (taste-dials roadmap)
 
 Within-project invariants + repetition caps. Override = declare the value in
 `DESIGN.md` (`## Consistency Locks` / a brutalist or uniform-grid style note).
 
-| # | Pattern | Why it reads as AI-generated | Override condition |
-|---|---|---|---|
-| V8 | Shape Lock — ≥ 4 distinct corner-radius scales in one surface | A fragmented radius system reads as no shape decision | Declare the radius scale in `DESIGN.md`; a genuine multi-tier scale documents each tier |
-| C6 | Colour Lock — ≥ 3 distinct saturated accent hue families | No single accent identity; reads as default-palette sprawl | Brand genuinely uses a multi-accent system, declared in `DESIGN.md` / `brand-to-tokens` |
-| C7 | Theme inversion — a light island mid-dark surface (or vice versa) not driven by the theme system | Mid-surface inversion breaks the theme rhythm | Intentional inverted callout with a documented reason — **judgment-only**, design-review calls it (deterministic detection is unreliable) |
-| L9 | Section monotony — ≥ 8 sections, < 4 distinct layout families | Uniform section rhythm collapses scannability | Deliberately uniform grid (e.g. a gallery), declared as such |
-| L10 | Zigzag — > 2 consecutive image-left/text-right two-column sections | The alternating-zigzag landing-page tell | A genuinely sequenced walkthrough where alternation aids comprehension |
+| # | Pattern | Why it reads as AI-generated | Override condition | Class | Remediation | Verification |
+|---|---|---|---|---|---|---|
+| V8 | Shape Lock — ≥ 4 distinct corner-radius scales in one surface | A fragmented radius system reads as no shape decision | Declare the radius scale in `DESIGN.md`; a genuine multi-tier scale documents each tier | invariant | reduce → retune | static |
+| C6 | Colour Lock — ≥ 3 distinct saturated accent hue families | No single accent identity; reads as default-palette sprawl | Brand genuinely uses a multi-accent system, declared in `DESIGN.md` / `brand-to-tokens` | invariant | reduce → retune | static |
+| C7 | Theme inversion — a light island mid-dark surface (or vice versa) not driven by the theme system | Mid-surface inversion breaks the theme rhythm | Intentional inverted callout with a documented reason — **judgment-only**, design-review calls it (deterministic detection is unreliable) | craft-presumption | delete → retune | judgment |
+| L9 | Section monotony — ≥ 8 sections, < 4 distinct layout families | Uniform section rhythm collapses scannability | Deliberately uniform grid (e.g. a gallery), declared as such | craft-presumption | retune | render |
+| L10 | Zigzag — > 2 consecutive image-left/text-right two-column sections | The alternating-zigzag landing-page tell | A genuinely sequenced walkthrough where alternation aids comprehension | craft-presumption | delete → retune | static |
 
 ---
 
 ## Visual
 
-| # | Pattern | Why it reads as AI-generated | Override condition |
-|---|---|---|---|
-| V1 | A colored accent border above 1px — the side stripe (`border-left` / `border-right`), a fully colored border on all four sides, or a gradient `border-image` | The single most recognizable AI-UI signature; every scaffold-default card uses it. Widened 2026-09-03: the side stripe was the 2024 form, the four-sided colored or gradient card border is the current one, and the entry as written reached only the first | Brand token explicitly defines the accent border; document the decision in the design brief. A 1px neutral border, a `border-color`-only focus treatment, and `transparent` / `currentColor` at any width are not accents and do not flag |
-| V2 | Glassmorphism as decoration (`backdrop-blur` + `bg-white/10`) with no functional depth hierarchy | Model defaults to it for "premium feel"; result looks like 2021 Material You clip-art | Element genuinely sits above a complex, blurred background layer; use sparingly and only for top-level floating surfaces |
-| V3 | Ghost card: `border: 1px solid` + box-shadow ≥ 16px spread together | The combination creates visual noise — the ghost card signals no confident decision about depth | Choose either border OR shadow to indicate depth; never both on the same surface |
-| V4 | Over-rounded small cards or badges: `border-radius` > 16px on elements < 200px wide | Excessive rounding on small surfaces looks toy-ish; standard cap is 12–16px | Pill shape (`border-radius: 9999px`) is intentional on single-line tags/chips only |
-| V5 | Hand-drawn or illustrated SVG icons mixed with crisp icon-system icons | Visual register collision; instantly reads as "generated asset" | Intentional illustrative section (hero illustration, empty-state art); isolate from the icon system |
-| V6 | Repeating diagonal stripes as a background texture | A cliché background default; reads as generated CSS art | Brand explicitly uses pattern backgrounds (e.g., a textile brand); document the stripe definition |
-| V7 | Nested cards: a card inside a card inside a card | Depth hierarchy collapses; user cannot parse elevation intent | A truly three-level hierarchy (e.g., project → board → card); ensure each level has distinct visual weight |
-| V9 | Noise or grain texture laid over a gradient as a "premium" surface default | The visual sibling of V6's diagonal stripes — a generated-CSS-art texture reached for to make a flat gradient look considered, applied because it is available rather than because the surface needed a texture | The texture is part of a documented brand surface treatment (a print-inspired or analog-photographic identity) with the grain opacity defined as a token, not chosen per surface |
+| # | Pattern | Why it reads as AI-generated | Override condition | Class | Remediation | Verification |
+|---|---|---|---|---|---|---|
+| V1 | A colored accent border above 1px — the side stripe (`border-left` / `border-right`), a fully colored border on all four sides, or a gradient `border-image` | The single most recognizable AI-UI signature; every scaffold-default card uses it. Widened 2026-09-03: the side stripe was the 2024 form, the four-sided colored or gradient card border is the current one, and the entry as written reached only the first | Brand token explicitly defines the accent border; document the decision in the design brief. A 1px neutral border, a `border-color`-only focus treatment, and `transparent` / `currentColor` at any width are not accents and do not flag | craft-presumption | delete → retune | static |
+| V2 | Glassmorphism as decoration (`backdrop-blur` + `bg-white/10`) with no functional depth hierarchy | Model defaults to it for "premium feel"; result looks like 2021 Material You clip-art | Element genuinely sits above a complex, blurred background layer; use sparingly and only for top-level floating surfaces | craft-presumption | delete → reduce | static |
+| V3 | Ghost card: `border: 1px solid` + box-shadow ≥ 16px spread together | The combination creates visual noise — the ghost card signals no confident decision about depth | Choose either border OR shadow to indicate depth; never both on the same surface | invariant | delete → reduce | static |
+| V4 | Over-rounded small cards or badges: `border-radius` > 16px on elements < 200px wide | Excessive rounding on small surfaces looks toy-ish; standard cap is 12–16px | Pill shape (`border-radius: 9999px`) is intentional on single-line tags/chips only | craft-presumption | retune | static |
+| V5 | Hand-drawn or illustrated SVG icons mixed with crisp icon-system icons | Visual register collision; instantly reads as "generated asset" | Intentional illustrative section (hero illustration, empty-state art); isolate from the icon system | invariant | replace | render |
+| V6 | Repeating diagonal stripes as a background texture | A cliché background default; reads as generated CSS art | Brand explicitly uses pattern backgrounds (e.g., a textile brand); document the stripe definition | style-preference | delete | static |
+| V7 | Nested cards: a card inside a card inside a card | Depth hierarchy collapses; user cannot parse elevation intent | A truly three-level hierarchy (e.g., project → board → card); ensure each level has distinct visual weight | invariant | delete → reduce | render |
+| V9 | Noise or grain texture laid over a gradient as a "premium" surface default | The visual sibling of V6's diagonal stripes — a generated-CSS-art texture reached for to make a flat gradient look considered, applied because it is available rather than because the surface needed a texture | The texture is part of a documented brand surface treatment (a print-inspired or analog-photographic identity) with the grain opacity defined as a token, not chosen per surface | style-preference | delete | render |
 
 ---
 
 ## Typography
 
-| # | Pattern | Why it reads as AI-generated | Override condition |
-|---|---|---|---|
-| T1 | Flat type hierarchy: body and headings use the same optical weight and nearly the same size | Type hierarchy is invisible; user cannot scan the page | Deliberately minimal aesthetic where all text is body-weight; must be an explicit brand decision |
-| T2 | Oversized italic serif as the primary hero headline | The "AI startup landing page" template; seen on ~60% of AI-generated marketing pages | Brand adopts editorial-magazine aesthetic with a documented italic serif rationale |
-| T3 | Icon tile (small rounded-square container) above heading | The universal AI feature-card template; reads as scaffold | Listed features genuinely benefit from visual-icon-then-title affordance AND the icons are custom, not library defaults |
-| T4 | Eyebrow / ALL-CAPS label above every section heading | Overuse collapses the emphasis; maximum is 1 eyebrow per 3 sections (countable: grep tracking classes) | Used for genuine section-type differentiation with meaningful vocabulary (not just "FEATURES", "PRICING") |
-| T5 | Repeated section kickers ("Our mission", "What we offer", "Why choose us") | Filler vocabulary; every AI-generated page has the same three section titles | Every kicker is substantive and uses brand-specific vocabulary, not generic marketing speak |
-| T6 | Crushed letter-spacing on display text (`letter-spacing` < −0.04em) | Below −0.04em glyphs collide optically; reads as "designer-ish effect" with no grounding | Intentional tight display treatment with a documented minimum floor; verify at the actual render size |
-| T7 | Overused fonts without a stated reason: Inter, Roboto, DM Sans, Geist, Space Grotesk, Instrument Serif | These are the default AI-coding-tool font picks; every generated UI uses one of them | **Register-scoped.** In the **brand** register: the brand has explicitly adopted one of these AND there is a documented reason. In the **product** register: a single reliable family is the sanctioned strategy ([`design-modes`](design-modes.md) § The two registers), so one of these is legitimate on a **stated** product-register choice — the register itself is the reason, and it must be stated (DESIGN.md, the surface brief, or `design-slop-disable`). Neither register admits an *undeclared* pick. |
-| T8 | Single typeface for everything | Missed typographic contrast opportunity; in the **brand** register an unexamined single family reads as "I did not think about fonts" | **Register-scoped like T7.** Intentional mono-typeface aesthetic (e.g., terminal-inspired), stated in the brief — *or* the **product** register, where one reliable family is the sanctioned strategy ([`design-modes`](design-modes.md) § The two registers) and contrast is carried by weight and scale rather than by a second family. |
-| T9 | All-caps body paragraphs or field labels | Reduces legibility below WCAG AA for sustained reading; reads as flair over function | Short UI labels (e.g., button text in a brutalist layout); never for body paragraphs |
-| T10 | Wide letter-spacing on body text (`letter-spacing` > 0.05em) | Reduces readability for sustained reading; reserved for display use | Display text (headlines, pull quotes) where intentional tracking is a brand element |
+| # | Pattern | Why it reads as AI-generated | Override condition | Class | Remediation | Verification |
+|---|---|---|---|---|---|---|
+| T1 | Flat type hierarchy: body and headings use the same optical weight and nearly the same size | Type hierarchy is invisible; user cannot scan the page | Deliberately minimal aesthetic where all text is body-weight; must be an explicit brand decision | craft-presumption | retune | render |
+| T2 | Oversized italic serif as the primary hero headline | The "AI startup landing page" template; seen on ~60% of AI-generated marketing pages | Brand adopts editorial-magazine aesthetic with a documented italic serif rationale | style-preference | replace | render |
+| T3 | Icon tile (small rounded-square container) above heading | The universal AI feature-card template; reads as scaffold | Listed features genuinely benefit from visual-icon-then-title affordance AND the icons are custom, not library defaults | style-preference | delete | static |
+| T4 | Eyebrow / ALL-CAPS label above every section heading | Overuse collapses the emphasis; maximum is 1 eyebrow per 3 sections (countable: grep tracking classes) | Used for genuine section-type differentiation with meaningful vocabulary (not just "FEATURES", "PRICING") | craft-presumption | delete → reduce | static |
+| T5 | Repeated section kickers ("Our mission", "What we offer", "Why choose us") | Filler vocabulary; every AI-generated page has the same three section titles | Every kicker is substantive and uses brand-specific vocabulary, not generic marketing speak | craft-presumption | replace | judgment |
+| T6 | Crushed letter-spacing on display text (`letter-spacing` < −0.04em) | Below −0.04em glyphs collide optically; reads as "designer-ish effect" with no grounding | Intentional tight display treatment with a documented minimum floor; verify at the actual render size | floor | retune | static |
+| T7 | Overused fonts without a stated reason: Inter, Roboto, DM Sans, Geist, Space Grotesk, Instrument Serif | These are the default AI-coding-tool font picks; every generated UI uses one of them | **Register-scoped.** In the **brand** register: the brand has explicitly adopted one of these AND there is a documented reason. In the **product** register: a single reliable family is the sanctioned strategy ([`design-modes`](design-modes.md) § The two registers), so one of these is legitimate on a **stated** product-register choice — the register itself is the reason, and it must be stated (DESIGN.md, the surface brief, or `design-slop-disable`). Neither register admits an *undeclared* pick. | reference-constraint | replace | judgment |
+| T8 | Single typeface for everything | Missed typographic contrast opportunity; in the **brand** register an unexamined single family reads as "I did not think about fonts" | **Register-scoped like T7.** Intentional mono-typeface aesthetic (e.g., terminal-inspired), stated in the brief — *or* the **product** register, where one reliable family is the sanctioned strategy ([`design-modes`](design-modes.md) § The two registers) and contrast is carried by weight and scale rather than by a second family. | reference-constraint | retune | judgment |
+| T9 | All-caps body paragraphs or field labels | Reduces legibility below WCAG AA for sustained reading; reads as flair over function | Short UI labels (e.g., button text in a brutalist layout); never for body paragraphs | floor | delete → retune | static |
+| T10 | Wide letter-spacing on body text (`letter-spacing` > 0.05em) | Reduces readability for sustained reading; reserved for display use | Display text (headlines, pull quotes) where intentional tracking is a brand element | craft-presumption | retune | static |
 
 ---
 
 ## Color
 
-| # | Pattern | Why it reads as AI-generated | Override condition |
-|---|---|---|---|
-| C1 | Purple/violet + cyan-on-dark palette as the primary scheme | The most recognizable AI-UI color combination in 2024–2026; instantly signals "generated" | Brand color system genuinely defines a violet primary; must come from `brand-to-tokens`, not from model defaults |
-| C2 | Gradient text via `background-clip: text` as a primary accent | Overused to the point of invisibility; also has rendering edge cases on some platforms | Headline treatment in a brand that explicitly uses gradient identity (e.g., a product whose logo is a gradient) |
-| C3 | Glowing colored accents on dark mode (bright neon `box-shadow` or `text-shadow`) | Looks like a 2020 gaming UI; reads as an AI attempt at "moody" | Brand explicitly defines neon accents with a documented usage rule (e.g., error or alert states in a security product) |
-| C4 | Gray text on colored background without WCAG AA check | Muted gray chosen for "elegance" often fails 4.5:1 contrast; the failure is invisible until audited | Any use — always verify the contrast ratio; if it fails, either darken the text or lighten/darken the background |
-| C5 | Cream/sand body background (OKLCH L 0.84–0.97, C < 0.06) with brass/clay/oxblood accents | The 2025 "premium-consumer" AI palette; every generated "sophisticated" product uses this exact combination | Brand has an explicit warm-neutral design language documented in `brand-to-tokens`; the palette is first-party, not a default reach |
+| # | Pattern | Why it reads as AI-generated | Override condition | Class | Remediation | Verification |
+|---|---|---|---|---|---|---|
+| C1 | Purple/violet + cyan-on-dark palette as the primary scheme | The most recognizable AI-UI color combination in 2024–2026; instantly signals "generated" | Brand color system genuinely defines a violet primary; must come from `brand-to-tokens`, not from model defaults | style-preference | replace | static |
+| C2 | Gradient text via `background-clip: text` as a primary accent | Overused to the point of invisibility; also has rendering edge cases on some platforms | Headline treatment in a brand that explicitly uses gradient identity (e.g., a product whose logo is a gradient) | style-preference | delete | static |
+| C3 | Glowing colored accents on dark mode (bright neon `box-shadow` or `text-shadow`) | Looks like a 2020 gaming UI; reads as an AI attempt at "moody" | Brand explicitly defines neon accents with a documented usage rule (e.g., error or alert states in a security product) | style-preference | delete → reduce | static |
+| C4 | Gray text on colored background without WCAG AA check | Muted gray chosen for "elegance" often fails 4.5:1 contrast; the failure is invisible until audited | Any use — always verify the contrast ratio; if it fails, either darken the text or lighten/darken the background | floor | retune | static |
+| C5 | Cream/sand body background (OKLCH L 0.84–0.97, C < 0.06) with brass/clay/oxblood accents | The 2025 "premium-consumer" AI palette; every generated "sophisticated" product uses this exact combination | Brand has an explicit warm-neutral design language documented in `brand-to-tokens`; the palette is first-party, not a default reach | style-preference | replace | static |
 
 ---
 
 ## Layout
 
-| # | Pattern | Why it reads as AI-generated | Override condition |
-|---|---|---|---|
-| L1 | Hero metric template: giant number, small label, 3–4 stats in a row, gradient | The default SaaS dashboard hero; appears on ~70% of AI-generated dashboard screens | Data storytelling genuinely requires a primary metric with supporting context; must vary the arrangement |
-| L2 | Three identical feature cards in a grid with the same visual weight | The scaffold-default; no visual rhythm, no hierarchy among features | Product genuinely has three equally weighted features; must still vary padding, or use a different composition |
-| L3 | Monotonous spacing: one spacing value (usually 24px) used for everything | Spacing is rhythm; uniform spacing collapses the hierarchy of proximity | Intentional grid where every element has equal weight; must document the grid definition |
-| L4 | Numbered section markers (01 / 02 / 03) as a primary section-order device | Universal AI-generated marketing page tell; seen on virtually every generated landing page | A genuinely numbered process or sequenced guide where the step number carries meaning beyond decoration |
-| L5 | Line length > 80ch in body copy without column constraint | Reduces readability; WCAG recommends 45–75 characters; beyond 80 the eye struggles to track | Data-dense dashboards where column width is driven by tabular data, not narrative text |
-| L6 | `position: absolute` dropdown inside `overflow: hidden` container | Dropdown clips; a common AI-generated layout bug | Refactor: use `<dialog>`, the Popover API, or `position: fixed` / portal to escape the stacking context |
-| L7 | Content that overflows its container | Text truncation or reflow failures at non-default viewport widths; reads as untested | Never acceptable in production output; test at 320px, 768px, 1280px minimally |
-| L8 | Z-index values of 99, 999, 9999 without a semantic z-index scale | Magic numbers; the first time two elements compete, the scale breaks | Establish a named semantic z-index scale (`--z-modal: 400`, `--z-toast: 500`, etc.) and use it |
+| # | Pattern | Why it reads as AI-generated | Override condition | Class | Remediation | Verification |
+|---|---|---|---|---|---|---|
+| L1 | Hero metric template: giant number, small label, 3–4 stats in a row, gradient | The default SaaS dashboard hero; appears on ~70% of AI-generated dashboard screens | Data storytelling genuinely requires a primary metric with supporting context; must vary the arrangement | craft-presumption | retune | render |
+| L2 | Three identical feature cards in a grid with the same visual weight | The scaffold-default; no visual rhythm, no hierarchy among features | Product genuinely has three equally weighted features; must still vary padding, or use a different composition | craft-presumption | retune | render |
+| L3 | Monotonous spacing: one spacing value (usually 24px) used for everything | Spacing is rhythm; uniform spacing collapses the hierarchy of proximity | Intentional grid where every element has equal weight; must document the grid definition | craft-presumption | retune | static |
+| L4 | Numbered section markers (01 / 02 / 03) as a primary section-order device | Universal AI-generated marketing page tell; seen on virtually every generated landing page | A genuinely numbered process or sequenced guide where the step number carries meaning beyond decoration | style-preference | delete | static |
+| L5 | Line length > 80ch in body copy without column constraint | Reduces readability; WCAG recommends 45–75 characters; beyond 80 the eye struggles to track | Data-dense dashboards where column width is driven by tabular data, not narrative text | floor | retune | render |
+| L6 | `position: absolute` dropdown inside `overflow: hidden` container | Dropdown clips; a common AI-generated layout bug | Refactor: use `<dialog>`, the Popover API, or `position: fixed` / portal to escape the stacking context | invariant | replace | render |
+| L7 | Content that overflows its container | Text truncation or reflow failures at non-default viewport widths; reads as untested | Never acceptable in production output; test at 320px, 768px, 1280px minimally | invariant | retune | render |
+| L8 | Z-index values of 99, 999, 9999 without a semantic z-index scale | Magic numbers; the first time two elements compete, the scale breaks | Establish a named semantic z-index scale (`--z-modal: 400`, `--z-toast: 500`, etc.) and use it | invariant | replace | static |
 
 ---
 
 ## Motion
 
-| # | Pattern | Why it reads as AI-generated | Override condition |
-|---|---|---|---|
-| M1 | Bounce or elastic easing curves on UI elements | Bounce draws attention to the animation itself; UI motion should serve the action, not perform | Intentional playful affordance (game UI, onboarding celebration state); must be brief and non-repeating |
-| M2 | Animating layout properties (width, height, top, left, padding) | Triggers browser layout recalculation; janky on mid-range devices; always solvable with `transform` | Never — use `transform: scaleX()` for width-like effects, `transform: translateY()` for position |
-| M3 | Animating `<img>` element on hover (scale, filter, transform) | Triggers composite + image decode; often causes content shift | Animate the card's background, border, or shadow instead; leave the image static |
-| M4 | `transition: all` shorthand | Animates every property including layout, color, opacity simultaneously; unpredictable and expensive | Enumerate only the properties that should animate: `transition: transform 200ms ease-out, opacity 200ms` |
-| M5 | Missing `@media (prefers-reduced-motion: reduce)` alternative for any animation | Accessibility violation; vestibular disorders make animated UI unusable | Never acceptable; always add a `prefers-reduced-motion` variant — gentler animation (reduced distance/opacity) is preferred over `display: none` |
-| M6 | Every section fading and rising into view on scroll | The most recognisable interaction tell of the current generation; a reveal applied per section carries no information — the reader has already scrolled to the content, so the animation delays what they asked for | The reveal carries meaning: a stepped narrative where each beat must land before the next, or a long-form article using position to signal progress. A reveal on *every* section is a default, not a decision |
-| M7 | A radial highlight tracking the pointer across a dark hero | A `mousemove` handler feeding a radial gradient; it draws the eye to the cursor, which is the one thing on the page the user already knows the position of | The spotlight is the interaction — a genuine reveal mechanic, a canvas inspection tool, an accessibility magnifier. Ambience is not a purpose |
-| M8 | An interactive control that reduces its own opacity on hover | Reads as the element retreating from the pointer rather than responding to it, and a fade is the one hover treatment that *reduces* legibility at the moment the user is committing to the target — see the six interaction states in [`design-review`](../../src/skills/design-review/SKILL.md) | The fade communicates a real state change (an item being dismissed, a layer being peeled back). Hover must stay legible; brightness, background or border carry hover without spending contrast |
+| # | Pattern | Why it reads as AI-generated | Override condition | Class | Remediation | Verification |
+|---|---|---|---|---|---|---|
+| M1 | Bounce or elastic easing curves on UI elements | Bounce draws attention to the animation itself; UI motion should serve the action, not perform | Intentional playful affordance (game UI, onboarding celebration state); must be brief and non-repeating | invariant | replace | feel |
+| M2 | Animating layout properties (width, height, top, left, padding) | Triggers browser layout recalculation; janky on mid-range devices; always solvable with `transform` | Never — use `transform: scaleX()` for width-like effects, `transform: translateY()` for position | floor | replace | static |
+| M3 | Animating `<img>` element on hover (scale, filter, transform) | Triggers composite + image decode; often causes content shift | Animate the card's background, border, or shadow instead; leave the image static | invariant | replace | render |
+| M4 | `transition: all` shorthand | Animates every property including layout, color, opacity simultaneously; unpredictable and expensive | Enumerate only the properties that should animate: `transition: transform 200ms ease-out, opacity 200ms` | invariant | replace | static |
+| M5 | Missing `@media (prefers-reduced-motion: reduce)` alternative for any animation | Accessibility violation; vestibular disorders make animated UI unusable | Never acceptable; always add a `prefers-reduced-motion` variant — gentler animation (reduced distance/opacity) is preferred over `display: none` | floor | delete → retune | static |
+| M6 | Every section fading and rising into view on scroll | The most recognisable interaction tell of the current generation; a reveal applied per section carries no information — the reader has already scrolled to the content, so the animation delays what they asked for | The reveal carries meaning: a stepped narrative where each beat must land before the next, or a long-form article using position to signal progress. A reveal on *every* section is a default, not a decision | craft-presumption | delete → reduce | feel |
+| M7 | A radial highlight tracking the pointer across a dark hero | A `mousemove` handler feeding a radial gradient; it draws the eye to the cursor, which is the one thing on the page the user already knows the position of | The spotlight is the interaction — a genuine reveal mechanic, a canvas inspection tool, an accessibility magnifier. Ambience is not a purpose | style-preference | delete | feel |
+| M8 | An interactive control that reduces its own opacity on hover | Reads as the element retreating from the pointer rather than responding to it, and a fade is the one hover treatment that *reduces* legibility at the moment the user is committing to the target — see the six interaction states in [`design-review`](../../src/skills/design-review/SKILL.md) | The fade communicates a real state change (an item being dismissed, a layer being peeled back). Hover must stay legible; brightness, background or border carry hover without spending contrast | craft-presumption | retune | feel |
 
 ---
 
 ## Copy
 
-| # | Pattern | Why it reads as AI-generated | Override condition |
-|---|---|---|---|
-| CP1 | Em-dash (`—`) overuse | The #1 syntactic LLM signature; appears in AI copy at 3–5× the frequency of human editorial copy; maximum 2 per 500 words | Specific editorial voice that uses em-dashes deliberately; count them before shipping |
-| CP2 | Marketing buzzwords: streamline, empower, supercharge, world-class, enterprise-grade, seamlessly, robust, leverage | These words carry no information; an AI-slop test: replace each with nothing — the sentence still communicates the same thing | No override; if a word can be deleted without loss, delete it |
-| CP3 | Aphoristic manufactured contrast ("We don't do X. We do Y." / "Not just A, but B.") | A generated sentence rhythm that sounds like brand writing but carries no specificity | Aphoristic copy is legitimate when the contrast reveals a genuine product distinction; test by asking whether the Y side says something concrete |
-| CP4 | "Theater" framing — copy that announces what the product does without saying how or why | "The platform for modern teams." communicates nothing; it is content-shaped noise | Every tagline must pass the "so what" test: can a skeptic follow it with "but how?" and get a concrete answer? |
-| CP5 | Emoji-decoration in UI markup — 🚀/✅/🎉 prepending headings, buttons, list items, or CTAs (`🚀 Get Started`) | Performative-enthusiasm coating; the default-startup-template tell. No emoji is better than decorative emoji | Brand/`DESIGN.md` declares a systematic emoji strategy, or the emoji is functional (status indicator, category marker tied to real meaning) |
-| CP6 | Stock render subject named instead of the product's own subject matter — "floating abstract shapes", "gradient mesh", "glowing orb", "particle field", "futuristic HUD" | These name the default thing a render becomes when nobody decided what it is about; same delete-test as CP2, applied to art direction — swap the phrase for what the product actually shows and the brief says more | The subject IS abstract by decision (a generative-art surface, a data-driven field whose data is named), and the brief says which decision and why |
+| # | Pattern | Why it reads as AI-generated | Override condition | Class | Remediation | Verification |
+|---|---|---|---|---|---|---|
+| CP1 | Em-dash (`—`) overuse | The #1 syntactic LLM signature; appears in AI copy at 3–5× the frequency of human editorial copy; maximum 2 per 500 words | Specific editorial voice that uses em-dashes deliberately; count them before shipping | craft-presumption | delete → reduce | static |
+| CP2 | Marketing buzzwords: streamline, empower, supercharge, world-class, enterprise-grade, seamlessly, robust, leverage | These words carry no information; an AI-slop test: replace each with nothing — the sentence still communicates the same thing | No override; if a word can be deleted without loss, delete it | craft-presumption | delete | static |
+| CP3 | Aphoristic manufactured contrast ("We don't do X. We do Y." / "Not just A, but B.") | A generated sentence rhythm that sounds like brand writing but carries no specificity | Aphoristic copy is legitimate when the contrast reveals a genuine product distinction; test by asking whether the Y side says something concrete | style-preference | delete → replace | judgment |
+| CP4 | "Theater" framing — copy that announces what the product does without saying how or why | "The platform for modern teams." communicates nothing; it is content-shaped noise | Every tagline must pass the "so what" test: can a skeptic follow it with "but how?" and get a concrete answer? | craft-presumption | replace | judgment |
+| CP5 | Emoji-decoration in UI markup — 🚀/✅/🎉 prepending headings, buttons, list items, or CTAs (`🚀 Get Started`) | Performative-enthusiasm coating; the default-startup-template tell. No emoji is better than decorative emoji | Brand/`DESIGN.md` declares a systematic emoji strategy, or the emoji is functional (status indicator, category marker tied to real meaning) | craft-presumption | delete | static |
+| CP6 | Stock render subject named instead of the product's own subject matter — "floating abstract shapes", "gradient mesh", "glowing orb", "particle field", "futuristic HUD" | These name the default thing a render becomes when nobody decided what it is about; same delete-test as CP2, applied to art direction — swap the phrase for what the product actually shows and the brief says more | The subject IS abstract by decision (a generative-art surface, a data-driven field whose data is named), and the brief says which decision and why | craft-presumption | replace | judgment |
 
 ---
 

package/package.json

@@ -1,7 +1,7 @@
 {
     "name": "@event4u/agent-config",
     "mcpName": "io.github.event4u-app/agent-config",
-    "version": "14.18.0",
+    "version": "14.19.0",
     "description": "Governed skills, rules and commands for AI coding tools (Claude Code, Cursor, Windsurf, Copilot) — every claim machine-checked, including the counts in these badges.",
     "license": "MIT",
     "private": false,

package/src/agent-src/README.md

@@ -17,7 +17,7 @@ it as `.augment/` (local projection).
 ├── contexts/       # Background context documents for the agent
 ├── guidelines/     # Coding guidelines by language (reference material)
 ├── rules/          # Always-active behavior rules (auto-loaded every conversation)
-├── skills/         # On-demand expertise (matched by topic)
+├── skills/         # On-demand expertise (named or trigger-matched)
 └── templates/      # File templates for roadmaps, features, skills, etc.
 ```
 

package/src/config/ci-local-parity.yml

@@ -174,6 +174,9 @@ ci_only:
   - id: check_finding_dispositions
     class: release
     reason: "adjudicates the release-PR self-review findings ledger, whose input JSON is produced by the self-review workflow job"
+  - id: report_skill_activation
+    class: release
+    reason: "declared for the reason `sweep_source_surfaces` is, and with the same nuance: it is not unreachable locally, it is outside the `task ci` closure on purpose. `src/config/release-gate-locality.yml` gives its release-validation job a `local:` command with `verify: true`, so `task release:verify` runs it before a release PR exists — and that is the only run that measures anything, because the transcript store the census reads is keyed on the directory it runs in and does not exist on a runner. Adding it to `task ci` would put an advisory report that gates on nothing into every local run"
   # road-to-source-silence Phase 5.3. Two modes, and the distinction is the whole
   # reason this is a declaration rather than a `task ci` entry.
   #
@@ -192,6 +195,9 @@ ci_only:
   # BEFORE the metadata is public, which is the only moment fixing it is free.
   # A contributor therefore discovers that failure locally by construction,
   # which is exactly the harm this checker exists to prevent.
+  - id: check_skill_activation_claim
+    class: release
+    reason: "the gating half of the same release job — it compares the published census figure against its record. Reachable locally through the same `task release:verify` row in src/config/release-gate-locality.yml, and outside the `task ci` closure for the same reason its sibling is"
   - id: sweep_source_surfaces
     class: network
     reason: "the CI invocation is the weekly five-surface census — needs `gh` for the PR surface and a full-depth clone for the ~9,000-commit trunk surface, writes counts not a verdict, and is deliberately non-failing; its GATE mode (`--gate-metadata`) is reachable locally on pre-push via src/scripts/hooks/prepush_metadata_sources.sh and by hand, so no failure is discoverable only after a push"

package/src/config/gate-coverage.yml

@@ -38,6 +38,18 @@
 #                pending = not yet emitting `scanned:`. The guard REPORTS these
 #                rather than skipping silently: its own partial coverage must be
 #                visible, or it repeats the sin it exists to catch.
+#   ci_invocation
+#                optional repo-relative workflow path this row's `argv` is pinned
+#                to. Rule 2 above was an instruction to the author and nothing
+#                more — the guard ran whatever `argv` said and never looked at
+#                how CI calls the gate, so a row could drift from its workflow,
+#                or the workflow step could be deleted, and this file would keep
+#                reporting healthy coverage for a gate CI no longer runs. Where
+#                a row names a workflow, that file must contain an invocation of
+#                the gate whose argument list equals `argv` exactly. Opt-in per
+#                row because most gates here have several CI invocations with
+#                different arguments, and one `argv` cannot be identical to all
+#                of them.
 #   canary       optional mutation recipe (Phase 7) — see CANARY below
 #
 # COVERAGE — THE HONEST DENOMINATOR (re-measured 2026-08-28)
@@ -52,6 +64,14 @@
 # CI-identical argv, and 75 of those 78 are listed below.** The manifest is
 # complete for every gate that can carry a floor, covering ~25 % of the population.
 #
+# RECOUNTED 2026-09-06, on the change that adds `lint_motion_authority_drift`.
+# Recomputed on this tree rather than incremented, because the block below
+# measured a tree that no longer exists: `count_gate_scripts()` returns **318**
+# and `grep -c '^  - id:'` returns **89** (88 before this row), so 89/318 =
+# 28.0 %. Five scripts and four rows arrived between 2026-09-03 and this change
+# without a recount in between — the same drift every block here keeps
+# describing.
+
 # RECOUNTED 2026-09-03, on the change that adds `check_agent_artifact_location`.
 # Recomputed on this tree rather than incremented: `count_gate_scripts()` returns
 # **313** and `grep -c '^  - id:'` returns **84** (83 before this row), so
@@ -380,6 +400,33 @@
 # contract is `dist == rewrite(src)`; `check_ci_local_parity`, whose corpus is
 # workflow files) carries no recipe and is reported UNPROVEN, never green.
 gates:
+  - id: lint_motion_authority_drift
+    argv: ["--quiet"]
+    min_scanned: 90
+    corpus: "every motion statement in the five carriers bound to `design-patterns.md` § Motion — 16 `motion.csv` rows, 84 `styles.csv` rows, 16 in-scope `design-rules-checklist.md` § 7 statements, 8 `ux-guidelines.csv` Animation rows and 4 `app-interface.csv` Animation rows: 128 at the time of writing. The floor is 90 rather than near 128 because `styles.csv` is 84 of the 128 and a style catalog legitimately loses rows; below 90 the scan has lost a whole carrier, which is the failure this floor exists to see. A `> 0` check could not: four of the five carriers could vanish and one 16-row CSV would still report success."
+    status: enforced
+    no_canary_reason: >-
+      A create-only canary cannot express this gate's violation. The corpus is
+      five NAMED files, so a planted file lands in no scanned root; every
+      violation it can have is a MODIFICATION — a duration widened past its
+      band, an easing family swapped, an allowlisted sentence reworded. Same
+      class as `lint_metric_consumers`. Discrimination is proven instead by the
+      gate's own `--self-test`: 13 cases, 8 rejecting, each driving the real CLI
+      against a throwaway five-carrier tree, and covering both directions of the
+      two override paths — an elastic curve refused without an override
+      condition and accepted with one, an overshoot refused in `styles.csv`
+      without a declared playful register and accepted with one.
+    note: >
+      road-to-one-motion-authority step 1.4. Six shipped carriers stated motion
+      timing and easing and disagreed; this gate makes § Motion the sole carrier
+      and checks the other five against it. It does not restate the authority's
+      numbers, it PARSES them, so the authority cannot drift away from its own
+      enforcement. HONEST LIMIT, stated in the gate's own docstring: the
+      authority bands five named interactions, and a dependent row naming an
+      interaction it does not band — scroll reveal, stagger list, parallax,
+      loading loop — is reported as `unbanded` and NOT failed, because failing it
+      would mean inventing a band the authority never stated. 12 rows are
+      unbanded today and the count is printed on the green path.
   - id: lint_metric_consumers
     argv: ["--quiet"]
     min_scanned: 5
@@ -687,6 +734,35 @@ gates:
       # canary. The gate also carries a `--self-test` with seven rejecting cases;
       # the two prove different things and neither replaces the other.
 
+  - id: check_finding_dispositions
+    argv: []
+    ci_invocation: .github/workflows/consistency.yml
+    min_scanned: 0
+    corpus: "recorded findings in the ledger for the CURRENT package version (agents/evidence/release-findings/<package.json version>.json)"
+    status: enforced
+    no_canary_reason: >
+      Both violations this gate detects are ABSENCES — a ledger file that is not
+      there, and an empty ledger missing its `no_findings_reason`. The canary
+      contract plants a file at a path that must not already exist, and neither
+      absence can be produced that way: an extra ledger planted under a canary
+      name changes nothing, because the bare invocation reads only
+      `<package.json version>.json`. The negative controls that do exist are in
+      `tests/scripts/check_finding_dispositions.test.ts` — present-empty-with-reason
+      exits 0, present-empty-without-reason exits 1, and the released-vs-unreleased
+      discriminator is pinned in both directions so the absent-ledger branch cannot
+      silently invert.
+    note: >
+      The floor is 0 and CANNOT be higher, which is the honest reading of this
+      row rather than a weakening of it. The corpus is one release's finding
+      count, and zero findings is a legitimate state — 14.17.0 and 14.18.0 both
+      legitimately record zero, because their self-review never completed. What
+      this row buys is therefore not a floor: it is `ci_invocation`. Before it,
+      the gate had exactly one caller, on a `release/*` head branch that is
+      deleted after the merge, and its removal would have been noticed by
+      nothing — which is how two releases shipped with no ledger. The row now
+      fails if `.github/workflows/consistency.yml` stops calling the gate, or
+      calls it with arguments this manifest does not reproduce.
+
   - id: check_review_schema
     argv: []
     min_scanned: 1
@@ -2531,3 +2607,97 @@ gates:
 
         - [ ] **1.1 A planted step.** Carries all three shape signals so the
               gate must report it from a directory that is not the roadmap root.
+
+  - id: lint_readme_size
+    argv: ["--quiet"]
+    ci_invocation: .github/workflows/consistency.yml
+    min_scanned: 1
+    corpus: >
+      exactly one file — README.md — so the floor is 1 and a scan of 0 means the
+      gate ran somewhere the file does not exist rather than that it found
+      nothing to say. The gate is registered here because the failure it exists
+      to end was not a wrong measurement but an unwatched one: an earlier
+      roadmap brought this file under control, archived complete on 2026-05-18,
+      and the file drifted back over the following three months while the only
+      guard — a line budget with 112 lines of headroom — stayed green through
+      every step. A budget that is not registered can be removed as quietly as
+      it was added, which is how the previous fix stopped holding.
+    no_canary_reason: >-
+      the corpus is ONE file at a fixed path, and the canary contract is
+      create-only. A planted `README.md` would have to overwrite the real one to
+      land in scope, so the plant is either impossible or destructive; every
+      violation this gate can have is a MODIFICATION of that single file — words
+      added, a fence pushed down, lines grown. Discrimination is proven by
+      `--self-test` instead (4 cases, 3 rejecting, one per budget), which is the
+      stronger control here: a single planted file could only exercise one of the
+      three budgets, and measuring one while two regress is precisely the failure
+      that made this gate insufficient in the first place.
+  - id: lint_reasoning_fixtures
+    argv: ["--quiet"]
+    min_scanned: 21
+    corpus: >
+      the 21 rows of tests/reasoning-layer-eval/trigger-fixtures.json. The floor
+      is the live count rather than a number below it because the corpus is a
+      committed fixture set that changes only by a deliberate edit — a reading
+      below 21 means a row was dropped, which is exactly what this gate should
+      refuse to pass over. The gate exists because the corpus had ZERO
+      validators: both were deleted in the Python retirement while five
+      surviving references still told a reader to run them, and nothing in
+      Taskfile.yml or .github/workflows/ called the corpus at all.
+    no_canary_reason: >-
+      the corpus is a single committed JSON file, and the canary contract is
+      create-only. A planted file lands in no row; every violation this gate can
+      have is a MODIFICATION of that file — a key added, a field dropped, an
+      enum value changed. Discrimination is proven by `--self-test` instead
+      (4 cases, 3 rejecting), which drives the real CLI over a fixture root for
+      each of the three malformation classes the roadmap names by hand: a sixth
+      key, a missing `discipline`, and `host: "medium"`.
+
+  - id: lint_skill_trigger_contradiction
+    argv: ["--quiet"]
+    min_scanned: 250
+    corpus: >
+      every SKILL.md under src/skills/ — 299 at the time of writing. The floor
+      sits at 250 rather than at 299 because the skill count moves with ordinary
+      work in both directions, while a reading far below it means the walk lost
+      the directory rather than that skills were deleted. The gate exists
+      because three skills shipped saying "explicit request only" in the
+      description a catalogue reads while publishing an "## Auto-trigger
+      keywords" section nothing in src/ parses — so the operational-looking half
+      was the inert one, and a reader met two opposite answers in one file.
+    no_canary_reason: >-
+      a create-only canary CAN plant a SKILL.md, but planting one that carries
+      both halves would prove only that the matcher fires, which the self-test
+      already proves over a fixture root — and the plant would land inside
+      src/skills/, a projected tree, where a stray file reddens unrelated
+      generator and count gates before this one is reached. Discrimination is
+      proven by `--self-test` (3 cases, 1 rejecting) covering both directions:
+      the contradicting pair reds, and each half ALONE stays green, which is
+      what stops a matcher that fired on the keyword section alone.
+
+  - id: check_skill_activation_claim
+    argv: []
+    ci_invocation: .github/workflows/release-validation.yml
+    min_scanned: 1
+    corpus: >
+      exactly one claim block — `skill-activation-census-zero` in
+      `docs/CLAIMS.md` — checked against the census record it was derived from,
+      so the floor is 1 and a scan of 0 means the claim was renamed or deleted
+      rather than that the gate found nothing to say. The gate deliberately does
+      NOT floor the census's own corpus: the report that reads it is
+      `report_skill_activation`, which is not a gate and must not become one, and
+      registering a `report_*` id here would fork the population
+      `_lib/gate_population.ts` classifies by prefix.
+    status: enforced
+    no_canary_reason: >-
+      the corpus is ONE claim block at a fixed id, and the canary contract is
+      create-only. A planted second `docs/CLAIMS.md` would have to overwrite the
+      real one to land in scope, and every violation this gate can have is a
+      MODIFICATION of that one block — a figure moved, a population redrawn, a
+      figure the sentence stops stating. Discrimination is proven by
+      `--self-test` instead (4 cases, 3 rejecting, one per class), and end to end
+      by hand: editing the published turn count, the human-named population count
+      and the `last_verified` date each exits 1, and restoring each exits 0.
+      What the ROW buys on top of that is `ci_invocation` — the census had a
+      Taskfile target and no caller at all, so its removal would have been
+      noticed by nothing.

package/src/config/gate-violation-baselines.json

@@ -34,9 +34,9 @@
       "note": "The nine active non-draft roadmaps that predate the `relates:` requirement (road-to-agentic-engineering-assurance, road-to-component-library-lifecycle, road-to-council-evidence-integrity, road-to-frontend-fidelity-calibration, road-to-monorepo-scope-and-detection, road-to-observability-plate, road-to-role-scoped-spawn-profiles, road-to-skill-delivery-over-mcp, road-to-standing-payload-diet). This is a NEW obligation rather than a repaired scan root, and the ratchet is the point: `depends:` shipped in the template with nothing requiring it and measured 0 adoption, so requiring `relates:` retroactively across the estate would be an owner-reserved write with no analysis behind the edges it would have to invent (held as step 4.3, deferred). Clearing an entry here is one roadmap at a time: run `agent-config roadmap:context` on it, write the rows the probe surfaces, and lower the count."
     },
     "check_source_size_budget": {
-      "count": 18068,
-      "landed": "2026-09-05",
-      "note": "Lowered 18,245 -> 18,230 on 2026-09-02 by drain/publication-integrity-option-a. Phase 2 Option A moved the authoring instruction out of the generator: release.ts, which sits ~530 lines PAST the 1500 cap so every line there costs one excess, loses the comment emission, a 16-line HEAD_DEFAULT doc block and its alias const, and gains a shorter writer docstring. The doc block moved to _lib/release_material.ts, which is 289 lines and therefore UNDER the cap, so the same prose now costs nothing. Net -15. The branch first came in at +8 and paid its own way down rather than raising the baseline.",
+      "count": 18062,
+      "landed": "2026-09-06",
+      "note": "Lowered 18,068 -> 18,062 on 2026-09-06 by drain/host-enforcement-truth. Collapsing the five per-host binding constants in install.ts into src/scripts/hooks/host_lowering.yaml removed 29 lines from a file ~3,900 past the 1500 cap, where every line costs one excess; the journal's surface column and its extracted migration helper gave 23 back. Lowered 18,245 -> 18,230 on 2026-09-02 by drain/publication-integrity-option-a. Phase 2 Option A moved the authoring instruction out of the generator: release.ts, which sits ~530 lines PAST the 1500 cap so every line there costs one excess, loses the comment emission, a 16-line HEAD_DEFAULT doc block and its alias const, and gains a shorter writer docstring. The doc block moved to _lib/release_material.ts, which is 289 lines and therefore UNDER the cap, so the same prose now costs nothing. Net -15. The branch first came in at +8 and paid its own way down rather than raising the baseline.",
       "reason": "Lowered 18,230 -> 18,209 on 2026-09-03 by drain/artifact-location. `doctor --strict` needed 16 lines of integration wiring in cmd_doctor.ts, which sits ~2,100 lines PAST the 1500 cap so every line there costs one excess. The branch first came in at +137 and paid its own way down rather than raising the baseline: the strict semantics, threshold, kill switch, payload helper and argv fragments moved to _cli/_doctor_strict.ts (~160 lines, under the cap, costs nothing), taking it to +15; then the 37-line drift renderer moved to _cli/_doctor_emit.ts. Net -21. The renderer was chosen deliberately: pure rendering, two dependencies, no argparse involvement and therefore none of the byte-parity risk the parser surface carries against the retired Python implementation; nothing outside cmd_doctor.ts imported it. An AI council on the residual split (a: raise by 15 / d: descope the feature); neither was taken, because a parallel branch demonstrated the same day that paying down is achievable in this codebase, which is the evidence the split lacked. This gate's own test comment prescribes it: a commit that splits a god-file lowers the excess and must carry the lowered baseline. MERGE RECONCILIATION 2026-09-03, 18,209 -> 18,205: drain/wired-instruments and drain/artifact-location each lowered from the SAME 18,230 base in parallel -- the branch to 18,226 (net -4, run_continuation_hook.ts's termination ladder moved to _lib/continuation_ladder.ts and runtime_journal.ts's prose to _lib/outcome_vocabularies.ts, both under the cap) and main to 18,209 (net -21, the doctor --strict split above). Taking either side alone is wrong in a different direction: main's value silently discards the branch's earned -4, and the branch's value RAISES the baseline against main and weakens the gate. The committed number is neither -- it is the gate's own output on the merged tree, where both reductions apply. THIS LOCAL READING IS ADMISSIBLE HERE, and the discriminator matters because this file's ci-parity:local-only entry records a 165 -> 164 lowering that was reverted for exactly the opposite case: that count depends on how the environment resolves the projection surface, whereas this gate walks only SOURCE_ROOTS = ['src'] (check_source_size_budget.ts:88) counting excess lines over the 1500 ceiling, and `git status --porcelain -uall -- src/` reported ZERO untracked files, so a clean CI checkout of this commit measures the same 18,205. If CI reads anything else, CI's number is the correct one and this note is the record of why. LOWERED 18,205 -> 18,193 on 2026-09-03 by fix/release-curation-ask. The change adds a fifth publication guard, and its own additions landed at +10 excess: release.ts 2,008 -> 2,018, every line of which is past the 1500 cap. Paid down rather than baselined, in the two ways this gate's doctrine prescribes. The guard's rationale (~13 lines of docstring, including why prMerged is a parameter rather than a call-site if) lives in release_publication.ts, 915 lines and therefore UNDER the cap, so the prose costs nothing; and the two pure text helpers _splitlines and _rstrip moved out of release.ts into release_env.ts (333 lines, also under the cap), taking release.ts to 1,996 -- 12 BELOW where it started. The branch came in at +10 and leaves at -12. Recorded while moving them: _splitlines is duplicated a third and fourth time at adoption_report.ts:106 and readme_linter.ts:418; collapsing all three onto one definition touches three modules and was left out of this change rather than done silently. LOWERED 18,193 -> 18,180 on 2026-09-03 by fix/release-branch-reuse. The step-1 branch decision -- twenty lines of if/else in release.ts -- became one call to checkout_release_branch in release_publication.ts, which is 1,019 lines and therefore UNDER the 1500 cap, so the twenty lines plus a ~35-line rationale now cost nothing. release.ts 1,996 -> 1,983, every one of those lines being past the cap. The drill grew 765 -> 918 for two new scenarios and the snapshot/restore that keeps them from writing the real CHANGELOG, and it is under the cap too. A change that adds a feature and lowers the excess is the shape this gate's doctrine asks for; the alternative was a +13 raise for logic that had no business in the god-file. LOWERED 18,180 -> 18,068 on 2026-09-05 by fix/release-gate-producer-parity. The change wires the governance-mix obligation into the writer and the two local guards and adds a release-gate locality registry; its own additions to release.ts landed at +19 excess, every one of those lines past the 1500 cap. Paid down rather than baselined, in the way this entry's doctrine prescribes: the 132-line test-count-trend unit (_TEST_COUNT_LINE_RE, _TEST_LIST_MAX_BUFFER, _count_from_list_result, _count_tests_current, _previous_test_count_from_changelog, _render_test_trend_line) moved from release.ts into release_publication.ts, which is under the cap and where it now belongs on merit rather than only on arithmetic -- the same diff makes the missing `Tests: N` footer a section-level publication blocker, so the writer that emits the footer and the guard that refuses its absence sit in one module. release.ts 1,983 -> 1,871. All four names are re-exported from release.ts, so tests/scripts/release.test.ts is unchanged. LOCAL READING, ADMISSIBLE ON THIS ENTRY'S OWN DISCRIMINATOR: the gate walks only SOURCE_ROOTS=['src'] counting excess over the ceiling, and `git status --porcelain -uall -- src/` reports ZERO untracked files with the two new src/ files staged, so a clean CI checkout of this commit measures the same 18,068. If CI reads anything else, CI's number is the correct one and this note is the record of why."
     },
     "ci-parity:local-only": {

package/src/config/release-gate-locality.yml

@@ -127,3 +127,15 @@ jobs:
     verify: false
     script: ~
     guard: ~
+
+  activation-census:
+    # The one row where the local command asserts MORE than the job it mirrors,
+    # not less, so `partial` does not apply: the transcript store the census
+    # reads is keyed on the directory it runs in and does not exist on a runner,
+    # so CI can only re-take an empty reading and check the published figure
+    # against its record. Run locally, the same command reads the real store,
+    # which is the only place the census means anything.
+    local: './scripts-run src/scripts/report_skill_activation && ./scripts-run src/scripts/check_skill_activation_claim'
+    verify: true
+    script: src/scripts/report_skill_activation.ts
+    guard: ~

package/src/scripts/_cli/cmd_doctor.ts

@@ -2881,7 +2881,7 @@ function _run_checks(
         'claude-command-wrappers': _check_claude_command_wrappers,
         'surface-state': _check_surface_state,
         'hook-wiring': _check_hook_wiring,
-        ...(runtime_wiring.wiringRunners({ packageRoot: _package_root(), iterOverrides: () => iter_setting_overrides({ cwd: project_root }) }) as unknown as Record<string, CheckRunner>),
+        ...(runtime_wiring.wiringRunners({ packageRoot: _package_root(), projectRoot: project_root, iterOverrides: () => iter_setting_overrides({ cwd: project_root }) }) as unknown as Record<string, CheckRunner>),
         ...(install_reach.reachRunners({ projectRoot: project_root, resolvableVersion: _current_package_version() }) as unknown as Record<string, CheckRunner>),
         'stale-orphans': _check_stale_orphans,
         overrides: () => _check_overrides(project_root),
@@ -2967,7 +2967,7 @@ function _run_checks_no_manifest(
         'claude-command-wrappers': _check_claude_command_wrappers,
         'surface-state': _check_surface_state,
         'hook-wiring': _check_hook_wiring,
-        ...(runtime_wiring.wiringRunners({ packageRoot: _package_root(), iterOverrides: () => iter_setting_overrides({ cwd: project_root }) }) as unknown as Record<string, CheckRunner>),
+        ...(runtime_wiring.wiringRunners({ packageRoot: _package_root(), projectRoot: project_root, iterOverrides: () => iter_setting_overrides({ cwd: project_root }) }) as unknown as Record<string, CheckRunner>),
         ...(install_reach.reachRunners({ projectRoot: project_root, resolvableVersion: _current_package_version() }) as unknown as Record<string, CheckRunner>),
         'stale-orphans': _check_stale_orphans,
         // BOTH registries need the id. Registering only the first one crashed with

package/src/scripts/_cli/cmd_handoff.ts

@@ -17,7 +17,7 @@
  *                      file (legacy copy-paste mode)
  *   --launch <host>    spawn a fresh interactive host session after
  *                      generating (claude: hook-seeded; codex: handoff as
- *                      initial prompt — no hook surface there)
+ *                      initial prompt — this package binds no hook there)
  *   --llm              narrative-polish seam; v1: clear not-implemented error
  *   --root <path>      repo root override (default: cwd)
  *

package/src/scripts/_cli/cmd_update.ts

@@ -446,7 +446,7 @@ export async function main(argv: string[] | null = null, options: MainOptions =
     const installed_version = options.installed_version || _detect_installed_version();
     const state_path = options.state_path || update_check.DEFAULT_STATE_PATH;
 
-    // AGENT_CONFIG_OFFLINE=1 (set by `install.py --offline`) is honored as an
+    // AGENT_CONFIG_OFFLINE=1 (set by `agent-config install --offline`) is honored as an
     // env-level kill-switch. Mirrors cmd_versions.py.
     const offline = args.offline || process.env['AGENT_CONFIG_OFFLINE'] === '1';
 

package/src/scripts/_cli/cmd_versions.ts

@@ -426,7 +426,7 @@ export function main(argv: string[] | null = null, options: MainOptions = {}): n
     const err = options.err ?? _stderrSink();
     const opts = _parse(argv ?? process.argv.slice(2), out, err);
 
-    // AGENT_CONFIG_OFFLINE=1 (set by `install.py --offline`) is honored as a
+    // AGENT_CONFIG_OFFLINE=1 (set by `agent-config install --offline`) is honored as a
     // global kill-switch even when the per-command --offline flag is absent.
     const offline = opts.offline || process.env['AGENT_CONFIG_OFFLINE'] === '1';