get-shit-pretty 0.6.2 → 0.7.0

package/gsp/skills/gsp-brand-refine/SKILL.md

@@ -1,5 +1,5 @@
 ---
-name: brand-refine
+name: gsp-brand-refine
 description: Targeted brand adjustments mid-project — tweak colors, typography, or spacing without re-running the full branding diamond
 user-invocable: true
 model: sonnet
@@ -13,21 +13,21 @@ allowed-tools:
   - WebFetch
 ---
 <context>
-You are a GSP brand refinement skill. You take targeted feedback about brand visual issues and surgically update token values — no need to re-run strategy or identity.
+You are a GSP brand refinement skill. You take targeted feedback about brand visual issues and surgically update the brand's `.yml` preset — no need to re-run strategy or identity.
 
-This skill modifies **tokens.json** and **palettes.json** only. If the user's feedback is strategic ("make the tone more playful") or narrative ("the brand story feels off"), redirect to `/gsp:brand-strategy` or `/gsp:brand-identity`.
+This skill modifies **`{brand-name}.yml`** — the single source of truth for brand tokens, patterns, constraints, effects, and intensity. If the user's feedback is strategic ("make the tone more playful") or narrative ("the brand story feels off"), redirect to `/gsp-brand-strategy` or `/gsp-brand-identity`.
 </context>
 
 <objective>
-Accept natural language feedback about brand visuals, identify which tokens are affected, apply targeted value changes, and log what changed.
+Accept natural language feedback about brand visuals, identify which `.yml` values are affected, apply targeted changes, and regenerate `STYLE.md` if it exists.
 
-**Input:** Natural language feedback (e.g., "accent is too muted", "heading weight too heavy")
-**Output:** Updated `tokens.json` and/or `palettes.json` + `REFINE-LOG.md`
+**Input:** Natural language feedback (e.g., "accent is too muted", "make buttons rounder", "more motion")
+**Output:** Updated `{brand-name}.yml` + regenerated `STYLE.md` (if exists) + `REFINE-LOG.md`
 **Agent:** None — inline skill, surgical edits
 </objective>
 
 <execution_context>
-@${CLAUDE_SKILL_DIR}/../../references/design-tokens.md
+@${CLAUDE_SKILL_DIR}/../gsp-brand-guidelines/design-tokens.md
 </execution_context>
 
 <rules>
@@ -43,33 +43,40 @@ Accept natural language feedback about brand visuals, identify which tokens are
 <process>
 ## Step 0: Locate brand and parse feedback
 
-Extract feedback from the user's input (everything after `/gsp:brand-refine`).
+Extract feedback from the user's input (everything after `/gsp-brand-refine`).
 
 If no feedback provided, use `AskUserQuestion`: "What would you like to adjust? (e.g., 'accent is too muted', 'heading font feels too heavy', 'spacing too tight')"
 
 Resolve brand from `.design/branding/` (one → use it, multiple → ask). Set `BRAND_PATH`.
-Check that `{BRAND_PATH}/patterns/tokens.json` exists. If not: "No tokens found. Run `/gsp:brand-patterns` first."
+Find the brand's `.yml` file: scan `{BRAND_PATH}/patterns/` for a `.yml` file that is NOT in `foundations/` or `components/`. If not found: "No brand style found. Run `/gsp-brand-guidelines` first."
+
+Also check if `{BRAND_PATH}/patterns/STYLE.md` exists (will need regeneration after changes).
 
 ## Step 1: Read current state
 
-Read `{BRAND_PATH}/patterns/tokens.json` once. Extract all sections relevant to the feedback:
+Read the brand `.yml` once. Map feedback to the relevant section:
 
-| Feedback signals | Token section |
+| Feedback signals | `.yml` section |
 |-----------------|---------------|
-| color, accent, muted, vibrant, contrast, tint, shade, hue | `color` |
-| font, heading, body, weight, size, line-height | `typography` |
-| spacing, padding, gap, tight, loose, dense | `spacing` |
-| shadow, elevation, depth, flat | `shadow` |
-| radius, rounded, sharp, corners | `borderRadius` |
-
-Also read `{BRAND_PATH}/identity/palettes.json` if it exists and feedback involves colors.
+| color, accent, muted, vibrant, contrast, tint, shade, hue | `tokens.color` + `dark_mode.color` |
+| font, heading, body, weight, size, line-height | `tokens.typography` |
+| spacing, padding, gap, tight, loose, dense | `tokens.spacing` |
+| shadow, elevation, depth, flat | `tokens.elevation` |
+| radius, rounded, sharp, corners | `tokens.shape` |
+| motion, speed, slow, fast, bounce, snap | `tokens.motion` + `intensity.motion` |
+| creative, playful, restrained, chaos, calm | `intensity.variance` |
+| dense, airy, packed, spacious | `intensity.density` |
+| button, card, input, badge, nav | `patterns.{component}` |
+| never, always, forbidden, required | `constraints` |
+| hover, click, press, animation, effect | `effects` |
+| dark mode, dark background, dark theme, night | `dark_mode.color` (+ any section above if the feedback applies to dark mode specifically) |
 
 ## Step 2: Propose changes
 
 Show a clear before/after for each affected token:
 
 ```
-  /gsp:brand-refine
+  /gsp-brand-refine
   ═══════════════════════════════════════
 
   Feedback: "the accent is too muted"
@@ -101,7 +108,7 @@ If regenerating, call the tints.dev API:
 ```
 WebFetch: https://www.tints.dev/api/{colorName}/{hexWithoutHash}
 ```
-Parse the response for the 11-stop OKLCH scale (50–950) and update `palettes.json`.
+Parse the response for the 11-stop OKLCH scale (50–950) and update `identity/palettes.json` as a reference artifact, then update the `.yml` color tokens from the new ramp.
 
 ### Typography changes involving scale ratio
 
@@ -115,8 +122,8 @@ Use `AskUserQuestion`:
 - **Cancel** — "keep current values"
 
 Apply confirmed changes:
-1. **tokens.json** — edit values in place with `Edit`. Preserve structure.
-2. **palettes.json** — update if palette was regenerated.
+1. **`{brand-name}.yml`** — edit values in place with `Edit`. Preserve structure.
+2. **`STYLE.md`** — if it exists, regenerate the affected sections (Patterns tables, Constraints lists, Effects tables, or Intensity dials) to reflect the `.yml` changes. Read the template from `${CLAUDE_SKILL_DIR}/../../templates/phases/style.md` for format reference.
 
 ## Step 4: Log and finish
 
@@ -134,7 +141,7 @@ Append to `{BRAND_PATH}/REFINE-LOG.md`:
 Display summary:
 
 ```
-  /gsp:brand-refine — {n} tokens updated
+  /gsp-brand-refine — {n} tokens updated
   ═══════════════════════════════════════
 
   Updated: {list of tokens}

package/gsp/skills/gsp-brand-research/SKILL.md

@@ -1,5 +1,5 @@
 ---
-name: brand-research
+name: gsp-brand-research
 description: Research your market and competitors
 user-invocable: true
 model: sonnet
@@ -27,7 +27,7 @@ Research market context that will inform brand strategy.
 
 <execution_context>
 @${CLAUDE_SKILL_DIR}/../../templates/phases/discover.md
-@${CLAUDE_SKILL_DIR}/../../references/design-trends.md (index only — agent loads specific trend files only after open research validates relevance)
+@${CLAUDE_SKILL_DIR}/design-trends.md (index only — agent loads specific trend files only after open research validates relevance)
 @${CLAUDE_SKILL_DIR}/../gsp-style/styles/INDEX.yml
 </execution_context>
 
@@ -43,7 +43,7 @@ Research market context that will inform brand strategy.
 
 Resolve brand from `.design/branding/` (one → use it, multiple → ask). Set `BRAND_PATH`.
 
-Read `{BRAND_PATH}/BRIEF.md`. If missing, tell user to run `/gsp:start` first.
+Read `{BRAND_PATH}/BRIEF.md`. If missing, tell user to run `/gsp-start` first.
 Read `{BRAND_PATH}/config.json` for `brand_mode`.
 
 ## Step 2: Confirm research scope
@@ -85,5 +85,5 @@ Update `{BRAND_PATH}/STATE.md`: set Phase 1 (Discover) to `complete`.
 
 ## Step 5: Phase transition
 
-Render phase transition (see `references/phase-transitions.md`).
+Invoke `/gsp-phase-transition` with phase `discover` and output directory `{BRAND_PATH}/research/`.
 </process>

package/gsp/{references → skills/gsp-brand-research}/design-trends.md

@@ -1,6 +1,6 @@
 # Design Trends Reference — Index
 
-Production-ready style presets and deep engineering references for current design trends. Consolidated in the `/gsp:style` skill.
+Production-ready style presets and deep engineering references for current design trends. Consolidated in the `/gsp-style` skill.
 
 Last updated: 2026-03-14
 
@@ -8,13 +8,13 @@ Last updated: 2026-03-14
 
 ## Style Presets (34)
 
-Style presets with structured tokens live in the `/gsp:style` skill at `skills/gsp-style/styles/`. Each preset is a YAML file with color, typography, shape, elevation, spacing, and motion tokens ready for W3C Design Token expansion.
+Style presets with structured tokens live in the `/gsp-style` skill at `skills/gsp-style/styles/`. Each preset is a YAML file with color, typography, shape, elevation, spacing, and motion tokens ready for W3C Design Token expansion.
 
-Run `/gsp:style --list` to see all available styles, or read `skills/gsp-style/styles/INDEX.yml` for the full catalog with tags and vibes.
+Run `/gsp-style --list` to see all available styles, or read `skills/gsp-style/styles/INDEX.yml` for the full catalog with tags and vibes.
 
 ## Deep Engineering References (9)
 
-Detailed CSS implementation guides, framework notes, and accessibility analysis live in the `/gsp:style` skill at `skills/gsp-style/trends/`:
+Detailed CSS implementation guides, framework notes, and accessibility analysis live in the `/gsp-style` skill at `skills/gsp-style/trends/`:
 
 | # | Trend | Description | Compat |
 |---|-------|-------------|--------|

package/gsp/skills/gsp-brand-strategy/SKILL.md

@@ -1,5 +1,5 @@
 ---
-name: brand-strategy
+name: gsp-brand-strategy
 description: Define positioning, voice, and messaging
 user-invocable: true
 model: opus
@@ -27,9 +27,9 @@ Define brand strategy and voice through interactive creative direction, then pro
 
 <execution_context>
 @${CLAUDE_SKILL_DIR}/../../templates/phases/strategy.md
-@${CLAUDE_SKILL_DIR}/../../references/brand-archetypes.md
-@${CLAUDE_SKILL_DIR}/../../references/positioning-frameworks.md
-@${CLAUDE_SKILL_DIR}/../../references/voice-tone.md
+@${CLAUDE_SKILL_DIR}/brand-archetypes.md
+@${CLAUDE_SKILL_DIR}/positioning-frameworks.md
+@${CLAUDE_SKILL_DIR}/voice-tone.md
 </execution_context>
 
 <rules>
@@ -45,7 +45,7 @@ Define brand strategy and voice through interactive creative direction, then pro
 ## Step 0: Resolve brand
 
 Resolve brand from `.design/branding/` (one → use it, multiple → ask). Set `BRAND_PATH`.
-If missing, tell user to run `/gsp:start` first.
+If missing, tell user to run `/gsp-start` first.
 
 ## Step 1: Load context
 
@@ -65,7 +65,7 @@ Frame as: "Here's where this brand can win." Keep it to 4-6 lines.
 
 ## Step 3: Archetype selection
 
-Read the personality direction from BRIEF.md (gathered during `/gsp:start`). Use it as the starting point — don't re-ask for personality. Deepen it into a structural archetype.
+Read the personality direction from BRIEF.md (gathered during `/gsp-start`). Use it as the starting point — don't re-ask for personality. Deepen it into a structural archetype.
 
 Use `AskUserQuestion` with 2-3 archetype candidates that align with the chosen personality direction. Each option:
 - **Label:** archetype name
@@ -146,5 +146,5 @@ If adjust → loop back. If confirmed → proceed.
 
 Update `{BRAND_PATH}/STATE.md`: set Phase 2 (Strategy) to `complete`.
 
-Render phase transition (see `references/phase-transitions.md`).
+Invoke `/gsp-phase-transition` with phase `strategy` and output directory `{BRAND_PATH}/strategy/`.
 </process>

package/gsp/skills/gsp-brand-sync/SKILL.md

@@ -1,5 +1,5 @@
 ---
-name: brand-sync
+name: gsp-brand-sync
 description: Sync brand to match a project's shipped state — tokens, voice, visual patterns, personality
 user-invocable: true
 model: opus
@@ -17,7 +17,7 @@ allowed-tools:
 <context>
 Standalone brand feedback skill. After a project ships (or mid-development), the codebase and content may have diverged from the brand system — adjusted colors, shifted tone of voice, evolved visual patterns. This skill detects those divergences across all brand dimensions and updates the brand to match.
 
-This is the standalone version of the feedback loop built into `/gsp:project-build` (build-time). Use this when:
+This is the standalone version of the feedback loop built into `/gsp-project-build` (build-time). Use this when:
 - A project evolved beyond its original brand during development
 - Manual tweaks were made post-build
 - The voice/tone landed differently than the strategy specified
@@ -33,8 +33,8 @@ Compare a project's shipped state against its source brand across all dimensions
 </objective>
 
 <execution_context>
-@${CLAUDE_SKILL_DIR}/../../references/design-tokens.md
-@${CLAUDE_SKILL_DIR}/../../references/chunk-format.md
+@${CLAUDE_SKILL_DIR}/../gsp-brand-guidelines/design-tokens.md
+@${CLAUDE_SKILL_DIR}/chunk-format.md
 </execution_context>
 
 <rules>
@@ -49,7 +49,7 @@ Compare a project's shipped state against its source brand across all dimensions
 ## Step 0: Resolve brand and project
 
 Resolve brand from `.design/branding/` (one → use it, multiple → ask). Set `BRAND_PATH`.
-Check that the brand has at least one of: `patterns/tokens.json`, `strategy/`, `identity/`. If none exist, tell the user: "No brand system found. Run `/gsp:brand-patterns` first."
+Check that the brand has at least one of: `patterns/{brand-name}.yml`, `strategy/`, `identity/`. If none exist, tell the user: "No brand system found. Run `/gsp-brand-guidelines` first."
 
 Verify the project codebase has shipped output — source files with components, copy, or styles.
 
@@ -60,7 +60,7 @@ mkdir -p {BRAND_PATH}/sync
 ```
 
 Spawn the `gsp-brand-syncer` agent with:
-- `BRAND_PATH` and all available brand files (tokens.json, strategy chunks, identity chunks, foundation chunks)
+- `BRAND_PATH` and all available brand files ({brand-name}.yml, strategy chunks, identity chunks, foundation chunks)
 - Project codebase location (working directory)
 - **Output path:** `{BRAND_PATH}/sync/`
 
@@ -74,7 +74,7 @@ Read `{BRAND_PATH}/sync/SYNC-REPORT.md`. Present a compact summary per dimension
 - **Tokens only** — just sync the quantitative token changes
 - **Pick by dimension** — choose which dimensions to sync
 - **Review each** — walk through every divergence individually
-- **Refine manually** — run `/gsp:brand-refine` to make targeted token adjustments instead
+- **Refine manually** — run `/gsp-brand-refine` to make targeted token adjustments instead
 - **Skip** — don't update the brand
 
 If "Pick by dimension", ask per dimension. If "Review each", walk through the Update Map from the report.
@@ -85,15 +85,15 @@ For "Removed" tokens: ask whether to remove from brand or keep (may be used by o
 
 Use the Update Map from the sync report. For each confirmed change:
 
-**Tokens** — edit `tokens.json` in place (preserve W3C structure). Update corresponding foundation chunks in `{BRAND_PATH}/patterns/foundations/` and style preset `.yml` if it exists.
+**Tokens** — edit the brand `.yml` preset in place. Regenerate affected STYLE.md sections if they exist.
 
 **Voice & tone** — update `{BRAND_PATH}/strategy/voice-and-tone.md` (adjust attributes, tone positions, style rules). Update `messaging.md` if messaging shifted.
 
-**Visual patterns** — update foundation chunks in `{BRAND_PATH}/patterns/foundations/`. Update component specs and identity chunks if visual identity evolved.
+**Visual patterns** — update the brand `.yml` patterns/constraints/effects blocks. Update component specs in `{BRAND_PATH}/patterns/components/` and identity chunks if visual identity evolved.
 
 **Personality** — update `{BRAND_PATH}/strategy/archetype.md` and `positioning.md`. Update `brand-platform.md` if values/promise shifted.
 
-Preserve chunk format per `references/chunk-format.md`. Update INDEX.md files if chunks were added.
+Preserve chunk format per `chunk-format.md`. Update INDEX.md files if chunks were added.
 
 ## Step 4: Summary
 

package/gsp/skills/gsp-brand-sync/chunk-format.md

@@ -0,0 +1,79 @@
+# Chunk Format Reference
+
+Standard format for all GSP phase output files. Chunks are the primary output — agents write chunks directly, not monoliths.
+
+## File Format
+
+Every chunk follows this structure:
+
+    # {Section/Component/Screen Name}
+
+    > Phase: {phase} | Brand/Project: {name} | Generated: {DATE}
+
+    ---
+
+    {Content for this chunk}
+
+    ---
+
+    ## Related
+
+    - [{Related chunk name}]({relative-path-to-related-chunk})
+
+## Naming Conventions
+
+- **Singular, kebab-case, lowercase:** "Buttons" → `button.md`, "Date Picker" → `date-picker.md`
+- **Screen files:** `screen-{NN}-{kebab-case-name}.md` (e.g., `screen-01-home.md`)
+
+## INDEX.md Format
+
+Each phase directory gets an INDEX.md manifest:
+
+    # {Phase Name}
+    > Phase: {phase} | Brand/Project: {name} | Generated: {DATE}
+
+    | Chunk | File | ~Lines |
+    |-------|------|--------|
+    | {Section} | [{filename}](./{filename}) | ~{N} |
+
+Lightweight — just a lookup table. No prose.
+
+## Rules
+
+- **Chunks are primary output** — agents write chunks directly to the phase directory
+- **No monoliths** — do not write a single large file then re-chunk it
+- **Size target:** 50-200 lines per chunk
+- **Self-contained:** each chunk must be understandable without loading other chunks
+- **Cross-references:** `## Related` section uses relative paths to related chunks
+- **Idempotent:** re-running a phase regenerates all chunks in that phase directory
+
+## Output Budgets
+
+Context is finite. Every line in a chunk is consumed by downstream agents. Budget accordingly.
+
+### Per-chunk budgets
+
+| Chunk type | Target | Hard max | Notes |
+|-----------|--------|----------|-------|
+| Phase chunk (design, research, etc.) | 50-150 lines | 200 lines | Self-contained, one concept per chunk |
+| INDEX.md | 10-30 lines | 50 lines | Lookup table only, no prose |
+| BUILD-LOG.md | 50-100 lines | 150 lines | Summary + tables, not narrative |
+| Component spec | 30-80 lines | 120 lines | Props, states, behavior — not full implementation |
+| Screen spec | 80-150 lines | 200 lines | Layout, components, interactions, states |
+
+### Per-phase budgets (total across all chunks)
+
+| Phase | Target total | Hard max | Typical chunks |
+|-------|-------------|----------|----------------|
+| Brief | 100-200 lines | 300 lines | 2-4 |
+| Research | 200-400 lines | 600 lines | 5-8 |
+| Design | 300-600 lines | 800 lines | 6-12 |
+| Critique | 100-200 lines | 300 lines | 2-4 |
+| Build log | 50-100 lines | 150 lines | 1 |
+| Review | 100-200 lines | 300 lines | 2-4 |
+
+### Terminal output (inline skills)
+
+- **Diagnostic** (doctor, progress): uncapped — user needs to see it, does not persist in agent context
+- **Greeting/status** (start): 20-40 lines
+- **Phase transitions**: 10-20 lines

package/gsp/skills/gsp-color/SKILL.md

@@ -0,0 +1,73 @@
+---
+name: gsp-color
+description: "Design color systems — palettes, contrast, semantic mapping, dark mode"
+user-invocable: true
+model: sonnet
+allowed-tools:
+  - Read
+  - Write
+  - AskUserQuestion
+  - Glob
+  - Grep
+  - WebFetch
+---
+<context>
+You are a GSP color director. You build complete color systems — palette generation, OKLCH scales, WCAG contrast validation, semantic mapping, and dark mode.
+
+This is a standalone composable skill. It works two ways:
+1. **Standalone** — user runs `/gsp-color` directly for palette exploration, contrast checking, or full system design
+2. **As a building block** — the creative-director invokes `/gsp-color --enrich` to add technical precision to creative color choices
+</context>
+
+<objective>
+Build production-ready color palettes or full color systems from brand colors or user input.
+
+**Input:** Hex colors, `--preview`, `--enrich`, or interactive
+**Output:** `color-system.md` chunk + `palettes.json` (OKLCH scales)
+**Agent:** None — inline skill, deterministic palette generation + contrast math
+</objective>
+
+<execution_context>
+@${CLAUDE_SKILL_DIR}/chunk-format.md
+</execution_context>
+
+<rules>
+- Always use `AskUserQuestion` for user interaction — never prompt via plain text
+- One decision per question — never batch multiple questions in a single message
+- Every palette gets the full 11-stop OKLCH scale: 50, 100, 200, 300, 400, 500, 600, 700, 800, 900, 950
+- Color names must be semantic (primary, secondary, accent, neutral) not literal (red, blue)
+- All foreground/background pairs must report WCAG AA contrast ratios
+- Dark mode mapping must maintain equivalent contrast relationships
+- Foundation chunks follow chunk-format.md format exactly
+</rules>
+
+<process>
+## Step 0: Parse mode
+
+| Input | Mode | Domain |
+|-------|------|--------|
+| `/gsp-color #hex [#hex...] --preview` | Preview scales | palette |
+| `/gsp-color #hex [#hex...]` | Generate from hex | palette |
+| `/gsp-color --enrich` | Enrich existing system | system |
+| `/gsp-color` | Interactive full system | system |
+
+## Step 1: Load domain
+
+Read the domain file for the detected mode:
+- **palette** mode → Read `${CLAUDE_SKILL_DIR}/domains/palette.md`
+- **system** mode → Read `${CLAUDE_SKILL_DIR}/domains/system.md`
+
+For system mode, also read `${CLAUDE_SKILL_DIR}/references/color-composition.md`.
+
+## Step 2: Execute domain framework
+
+Follow the loaded domain file's complete workflow — it contains all generation logic, API calls, output formats, and completion steps.
+
+## Step 3: Write output
+
+Write `color-system.md` + `palettes.json` to the resolved output path (skip if `--preview`).
+
+## Step 4: Summary
+
+Display result summary and offer next steps per the domain file's completion section.
+</process>

package/gsp/skills/gsp-color/chunk-format.md

@@ -0,0 +1,79 @@
+# Chunk Format Reference
+
+Standard format for all GSP phase output files. Chunks are the primary output — agents write chunks directly, not monoliths.
+
+## File Format
+
+Every chunk follows this structure:
+
+    # {Section/Component/Screen Name}
+
+    > Phase: {phase} | Brand/Project: {name} | Generated: {DATE}
+
+    ---
+
+    {Content for this chunk}
+
+    ---
+
+    ## Related
+
+    - [{Related chunk name}]({relative-path-to-related-chunk})
+
+## Naming Conventions
+
+- **Singular, kebab-case, lowercase:** "Buttons" → `button.md`, "Date Picker" → `date-picker.md`
+- **Screen files:** `screen-{NN}-{kebab-case-name}.md` (e.g., `screen-01-home.md`)
+
+## INDEX.md Format
+
+Each phase directory gets an INDEX.md manifest:
+
+    # {Phase Name}
+    > Phase: {phase} | Brand/Project: {name} | Generated: {DATE}
+
+    | Chunk | File | ~Lines |
+    |-------|------|--------|
+    | {Section} | [{filename}](./{filename}) | ~{N} |
+
+Lightweight — just a lookup table. No prose.
+
+## Rules
+
+- **Chunks are primary output** — agents write chunks directly to the phase directory
+- **No monoliths** — do not write a single large file then re-chunk it
+- **Size target:** 50-200 lines per chunk
+- **Self-contained:** each chunk must be understandable without loading other chunks
+- **Cross-references:** `## Related` section uses relative paths to related chunks
+- **Idempotent:** re-running a phase regenerates all chunks in that phase directory
+
+## Output Budgets
+
+Context is finite. Every line in a chunk is consumed by downstream agents. Budget accordingly.
+
+### Per-chunk budgets
+
+| Chunk type | Target | Hard max | Notes |
+|-----------|--------|----------|-------|
+| Phase chunk (design, research, etc.) | 50-150 lines | 200 lines | Self-contained, one concept per chunk |
+| INDEX.md | 10-30 lines | 50 lines | Lookup table only, no prose |
+| BUILD-LOG.md | 50-100 lines | 150 lines | Summary + tables, not narrative |
+| Component spec | 30-80 lines | 120 lines | Props, states, behavior — not full implementation |
+| Screen spec | 80-150 lines | 200 lines | Layout, components, interactions, states |
+
+### Per-phase budgets (total across all chunks)
+
+| Phase | Target total | Hard max | Typical chunks |
+|-------|-------------|----------|----------------|
+| Brief | 100-200 lines | 300 lines | 2-4 |
+| Research | 200-400 lines | 600 lines | 5-8 |
+| Design | 300-600 lines | 800 lines | 6-12 |
+| Critique | 100-200 lines | 300 lines | 2-4 |
+| Build log | 50-100 lines | 150 lines | 1 |
+| Review | 100-200 lines | 300 lines | 2-4 |
+
+### Terminal output (inline skills)
+
+- **Diagnostic** (doctor, progress): uncapped — user needs to see it, does not persist in agent context
+- **Greeting/status** (start): 20-40 lines
+- **Phase transitions**: 10-20 lines