Skill v1.0.1

version: "1.0.1" name: gsp-project-brief description: Scope a feature or flow — use when: scope this, define what we're building, plan the project, what should we build, write a brief user-invocable: true allowed-tools:

• Read
• Write
• Bash
• Grep
• Glob

<context> Phase 1 of the GSP project diamond. Scopes the project by determining what screens and components are needed, what adaptations the brand system requires for this specific project, and performs gap analysis against the codebase.

Encourages treating the project as a bounded issue (or set of issues) and a PR — ship small, ship complete.

Works with the dual-diamond architecture: reads brand system from .design/branding/{brand}/patterns/ via brand.ref, reads/writes project assets in .design/projects/{project}/. </context>

<objective> Scope the project and plan adaptations from the brand system.

Input: Brand system (via brand.ref) + project BRIEF.md + config.json Output: {project}/brief/ (scope.md, target-adaptations.md, conditionals, INDEX.md) Output mode: Honor ${CLAUDE_SKILL_DIR}/../../policies/output-modes.md per preferences.project_size (default compact). Chunk file counts and consolidation rules live in the policy. </objective>

<execution_context> @${CLAUDE_SKILL_DIR}/../../templates/phases/brief.md </execution_context>

<process>

Step 0: Resolve project and brand

If .design/projects/ does not exist: output "No GSP project found. Run /gsp-start to begin." and stop.

Resolve project from .design/projects/ (one → use it, multiple → ask). Set PROJECT_PATH.

Read {PROJECT_PATH}/brand.ref → set BRAND_PATH. If brand.ref doesn't exist, tell the user to run /gsp-start.

Step 1: Load context

Brand patterns (chunk-first)

Read {BRAND_PATH}/patterns/INDEX.md. If it exists, load all foundation chunks + selective component chunks.

If it doesn't exist, tell the user to run /gsp-brand-guidelines first to create the brand's design patterns.

Also read the brand .yml preset from {BRAND_PATH}/patterns/.

Project context

Read:

• {PROJECT_PATH}/BRIEF.md — what we're building, platforms, tech stack
• {PROJECT_PATH}/config.json — get implementation_target, design_scope, codebase_type, app_path, repo_type, accessibility_level, style_preset

After reading config, check if app_path is empty. If empty AND (repo_type is monorepo OR multiple package.json files are found at apps/*/package.json or packages/*/package.json), set flag NEEDS_APP_SELECTION = true.

Expertise pointers:

• accessibility_level (AA vs AAA) carries scope implications for contrast-sensitive components — see ${CLAUDE_SKILL_DIR}/../gsp-accessibility/SKILL.md for level interpretation when reasoning about adaptations
• style_preset (when set) drives style-preset-specific component conventions — see ${CLAUDE_SKILL_DIR}/../gsp-style/styles/{preset}.md for the preset's patterns + constraints + effects vocabulary that adaptation reasoning should respect
• When writing token overrides in target-adaptations.md, consult ${CLAUDE_SKILL_DIR}/../gsp-color/domains/palette.md for OKLCH/contrast considerations and ${CLAUDE_SKILL_DIR}/../gsp-typography/domains/scale.md for type-scale ratios

Codebase context

Read .design/system/STACK.md and .design/system/COMPONENTS.md (if exist) — existing tech stack, components, architecture.

Stack inheritance (existing codebases): If STACK.md exists and the project config.json has empty or generic values for tech_stack, implementation_target, or codebase_type, inherit them from STACK.md directly — do not ask the user questions the stack already answers. Update config.json with the inherited values before proceeding.

The global stack is the compliance baseline. Every project in this workspace must target it. If the user's brief describes a different stack than what STACK.md declares (e.g., "I want to use Radix directly" when STACK.md says shadcn/ui), surface the conflict: "⚠️ Your brief mentions {X}, but the workspace stack is {Y} per STACK.md. Proceed with the workspace stack, or update STACK.md first?"

Read .design/CHANGELOG.md — quick history of what prior projects built. For projects with overlapping scope, read their codebase/MANIFEST.md for detail. Glob .design/projects/*/STATE.md — detect active sibling projects.

Step 1.3: App targeting

Run when NEEDS_APP_SELECTION = true.

1. Glob for apps/*/package.json and packages/*/package.json to find all app packages.
2. Read each package.json to extract name and the primary framework dependency (Next.js, Vite, React Native, etc.).
3. Build an app list, e.g.:

`` apps/web → my-app-web (Next.js) apps/mobile → my-app-mobile (Expo / React Native) packages/ui → my-app-ui (Vite) ``

1. Use AskUserQuestion: "Which app does this project target?" with one option per detected app (showing path + framework) plus Whole repo (no specific app).
2. Based on the user's choice:

• If an app was selected: set app_path to the chosen path (e.g. apps/web) and repo_type to monorepo
• If "Whole repo" was selected: set app_path to "" and repo_type to monorepo

1. Write the updated app_path and repo_type back to {PROJECT_PATH}/config.json.
2. Derive APP_NAME = last segment of app_path (e.g. apps/web → web, empty → root). Note that the per-app stack file will live at .design/system/stacks/{APP_NAME}.md.

Step 1.5: Scope check

If `design_scope` is `tokens`:

1. Update {PROJECT_PATH}/STATE.md — set Phase 1 (Brief) status to skipped
2. Display: "Brief phase skipped — design scope is tokens."
3. Route: "Run /gsp-project-build."
4. Stop here.

Step 1.7: Issue framing

Ask the user (via AskUserQuestion): "Is this project linked to a GitHub issue?"

• Yes — paste the issue URL → read the issue number from the URL, write git.issue to {PROJECT_PATH}/config.json and populate the | Issue | row in STATE.md. Also read the issue title/body via gh issue view {number} --json title,body,labels and use that content to pre-populate scope context (saves the user from re-describing what the issue already states).
• No — describe it → continue with manual brief as before
• Skip → continue without issue link

Step 2: Scope the project

Using all context loaded in Step 1, scope the project directly. Act as a Senior Design Project Lead bridging the brand system and the project's specific needs.

If any sibling project is active and its scope overlaps with this project, flag it: "⚠️ {name} is actively working on {scope}. Coordinate to avoid conflicts." If this project modifies components from a sibling's manifest, note provenance.

Scoping process

1. Analyze brief — what's being built, for whom, on what platforms
2. Define screen list — prioritized screens from brief, user flows, success criteria
3. Map component scope — which brand system components this project needs
4. Identify adaptations — project-specific variants, overrides, or extensions to brand components
5. Map to implementation target — connect design components to target primitives (shadcn, rn-reusables, existing, code)
6. Gap analysis (existing codebases) — what's in the brand system but missing from the codebase
7. Generate install manifest (shadcn/rn-reusables) — install commands for needed components
8. Issue framing — suggest how to break the project into bounded, shippable issues

Quality standards

• Every screen has a clear purpose and priority level
• Component adaptations reference specific brand system components
• Gap analysis is concrete (component names, token names)
• Install manifests are copy-paste ready
• Scope boundaries are explicit (what's in, what's out)

Write chunks to {PROJECT_PATH}/brief/

Use the brief output template from execution_context for chunk formatting.

1. `scope.md` (~80-120 lines) — prioritized screen list, component scope, project boundaries, success criteria, dependencies, issue framing
2. `target-adaptations.md` (~60-100 lines) — token overrides, component adaptations, platform considerations, implementation target mapping
3. `install-manifest.md` (shadcn/rn-reusables only) — install commands for all needed components
4. `gap-analysis.md` (existing target only) — components/tokens in brand system but not in codebase
5. `file-references.md` (existing target only) — paths to existing components/tokens being used

Cross-references: target-adaptations.md links to {BRAND_PATH}/patterns/components/{name}.md; gap-analysis.md links to brand system components and tokens; scope.md references the project BRIEF.md.

Write INDEX.md

# Brief
> Phase: brief | Project: {name} | Generated: {DATE}
 
## Scoping
 
| Chunk | File | ~Lines |
|-------|------|--------|
| Scope | [scope.md](./scope.md) | ~{N} |
| Target Adaptations | [target-adaptations.md](./target-adaptations.md) | ~{N} |
| Install Manifest | [install-manifest.md](./install-manifest.md) | ~{N} |
| Gap Analysis | [gap-analysis.md](./gap-analysis.md) | ~{N} |
| File References | [file-references.md](./file-references.md) | ~{N} |

Only include rows for chunks that were actually produced.

Step 3: Write exports

Update {PROJECT_PATH}/exports/INDEX.md:

• If INDEX.md doesn't exist, copy it from templates/exports-index.md
• Replace everything between <!-- BEGIN:brief --> and <!-- END:brief --> with populated table:

<!-- BEGIN:brief -->
| Section | File |
|---------|------|
| Scope | [scope.md](../brief/scope.md) |
| Target Adaptations | [target-adaptations.md](../brief/target-adaptations.md) |
| Install Manifest | [install-manifest.md](../brief/install-manifest.md) |
| Gap Analysis | [gap-analysis.md](../brief/gap-analysis.md) |
| File References | [file-references.md](../brief/file-references.md) |
<!-- END:brief -->

Only include rows for chunks that were actually produced.

Step 4: Update state

Update {PROJECT_PATH}/STATE.md:

• Set Phase 1 (Brief) status to complete
• Record completion date

Write/update .design/CLAUDE.md — register the project as started. If the file doesn't exist, read the template from ${CLAUDE_SKILL_DIR}/../../templates/design-claude.md first. Append under ## Projects:

### {project-name} · in progress · {DATE}
brand: {brand-name} · next: gsp-project-research · .design/projects/{project-name}/

Step 5: Phase transition output

Invoke /gsp-phase-transition with phase brief and output directory {PROJECT_PATH}/brief/. </process> </output>