bsmnt 0.1.2 → 0.1.3

package/docs/sanity-setup-steps.md

@@ -1,199 +0,0 @@
-# Sanity Project Setup — Automated Steps
-
-> Reproducible steps for programmatically creating a Sanity project, dataset, API token, and `.env.local` configuration. No browser or interactive prompts required.
-
----
-
-## Prerequisites
-
-- `@sanity/cli` available via `npx` (comes with `sanity` package)
-- User must be logged in to Sanity CLI (`npx sanity login`)
-- Project already has `next-sanity`, `@sanity/client`, `sanity` installed
-
----
-
-## Step 1: Verify Sanity CLI is available
-
-```bash
-npx sanity --version
-```
-
-If this fails, the sanity packages aren't installed. Run:
-
-```bash
-bun add next-sanity @sanity/client @sanity/image-url sanity @sanity/vision
-```
-
----
-
-## Step 2: Verify user is authenticated
-
-```bash
-npx sanity projects list
-```
-
-- If this returns a list of projects, the user is logged in.
-- If it fails with an auth error, the user needs to run `npx sanity login` first (this is interactive and opens a browser — cannot be automated).
-
----
-
-## Step 3: Create the Sanity project
-
-```bash
-npx sanity projects create --name "<project-name>" --dataset production
-```
-
-**Output to parse:**
-
-```
-Project created successfully!
-ID: <project-id>
-Name: <project-name>
-Organization: <org>
-Dataset: production (public)
-```
-
-**Extract the project ID** from the output. It's the alphanumeric string after `ID: ` (e.g., `1poovu2i`).
-
----
-
-## Step 4: Get the user's auth token
-
-The Sanity CLI stores the user's session token locally. Retrieve it with:
-
-```bash
-npx sanity debug --secrets 2>&1 | grep "Auth token"
-```
-
-**Output:**
-
-```
-  Auth token: '<token-string>'
-```
-
-**Parse the token** from between the single quotes. Note: the output includes ANSI color codes (`[32m` and `[39m`) that need to be stripped.
-
----
-
-## Step 5: Create an API read token
-
-Use the Sanity HTTP API to create a viewer token for the project:
-
-```bash
-curl -s -X POST "https://api.sanity.io/v2021-06-07/projects/<project-id>/tokens" \
-  -H "Authorization: Bearer <auth-token-from-step-4>" \
-  -H "Content-Type: application/json" \
-  -d '{"label": "Next.js Read Token", "roleName": "viewer"}'
-```
-
-**Response:**
-
-```json
-{
-  "id": "si8WgRxQuts4CN",
-  "key": "<the-api-read-token>",
-  "roles": [{ "name": "viewer", "title": "Viewer" }],
-  "label": "Next.js Read Token"
-}
-```
-
-**Extract the `key` field** — this is the `SANITY_API_READ_TOKEN`.
-
----
-
-## Step 6: Write `.env.local`
-
-Create `.env.local` in the project root with:
-
-```env
-# Sanity CMS
-NEXT_PUBLIC_SANITY_PROJECT_ID=<project-id>
-NEXT_PUBLIC_SANITY_DATASET=production
-SANITY_API_READ_TOKEN=<api-read-token>
-NEXT_PUBLIC_SANITY_API_READ_TOKEN=<api-read-token>
-```
-
-**Why both `SANITY_API_READ_TOKEN` and `NEXT_PUBLIC_SANITY_API_READ_TOKEN`?**
-
-The next-starter template's `env.ts` reads from `NEXT_PUBLIC_SANITY_API_READ_TOKEN` (exposed to client for draft mode / visual editing). `SANITY_API_READ_TOKEN` is the conventional server-only name. Setting both ensures compatibility regardless of which convention the project follows.
-
----
-
-## Step 7: Verify the connection
-
-Test that the token and project ID work by querying the API:
-
-```bash
-curl -s "https://<project-id>.api.sanity.io/v2024-03-15/data/query/production?query=*%5B0%5D" \
-  -H "Authorization: Bearer <api-read-token>"
-```
-
-**Success response:**
-
-```json
-{
-  "query": "*[0]",
-  "result": { ... },
-  "ms": 2
-}
-```
-
-If the response contains `"result"` (even if it's `null` for an empty dataset), the setup is correct.
-
----
-
-## Step 8: Ensure `.env.local` is gitignored
-
-Check if `.gitignore` exists and includes `.env.local`. If not, add it:
-
-```bash
-# Check
-grep -q ".env.local" .gitignore 2>/dev/null
-
-# If missing or .gitignore doesn't exist, ensure it's added
-echo ".env.local" >> .gitignore
-```
-
----
-
-## Summary of values produced
-
-| Variable | Source |
-|----------|--------|
-| `NEXT_PUBLIC_SANITY_PROJECT_ID` | Step 3 — `ID` field from project creation output |
-| `NEXT_PUBLIC_SANITY_DATASET` | Hardcoded to `production` (passed in Step 3) |
-| `SANITY_API_READ_TOKEN` | Step 5 — `key` field from token creation response |
-| `NEXT_PUBLIC_SANITY_API_READ_TOKEN` | Same as above |
-
----
-
-## Error handling notes
-
-| Step | Possible failure | How to handle |
-|------|-----------------|---------------|
-| 2 | Not logged in | Prompt user to run `npx sanity login` — requires browser |
-| 3 | Org selection needed | Add `--organization <org-id>` flag if user has multiple orgs |
-| 4 | Token not found | Auth token location varies by OS. Fallback: read `~/.config/sanity/config.json` |
-| 5 | 401/403 from API | Auth token expired or invalid — user needs to `npx sanity login` again |
-| 5 | 400 bad request | Check project ID is correct, check auth header format |
-| 7 | Connection refused | Project ID or dataset name is wrong |
-
----
-
-## CLI integration notes
-
-For a `basement init` command, the flow would be:
-
-```
-1. Check: `npx sanity projects list` → confirms auth
-2. Prompt: "Project name?" (default: directory name)
-3. Run: `npx sanity projects create --name <name> --dataset production`
-4. Parse: project ID from stdout
-5. Get: auth token from `npx sanity debug --secrets`
-6. Create: API token via HTTP POST
-7. Write: `.env.local` with all four variables
-8. Verify: curl test query
-9. Print: "Sanity Studio available at /studio"
-```
-
-No interactive Sanity prompts needed. The only prerequisite is that `npx sanity login` has been run at least once.

package/packages/cli/package.json

@@ -1,16 +0,0 @@
-{
-	"name": "@basementstudio/cli",
-	"version": "0.2.2",
-	"description": "CLI to scaffold basement projects and add integrations",
-	"type": "module",
-	"bin": {
-		"basement": "./bin/index.js"
-	},
-	"dependencies": {
-		"@basementstudio/create-basement-app": "workspace:*",
-		"fs-extra": "^11.3.3",
-		"ora": "^9.1.0",
-		"picocolors": "^1.1.1",
-		"prompts": "^2.4.2"
-	}
-}

package/tasks/.last-branch

@@ -1 +0,0 @@
-BertovDev/project-restructure

package/tasks/CLAUDE.md

@@ -1,104 +0,0 @@
-# Ralph Agent Instructions
-
-You are an autonomous coding agent working on a software project.
-
-## Your Task
-
-1. Read the PRD at `prd.json` (in the same directory as this file)
-2. Read the progress log at `progress.txt` (check Codebase Patterns section first)
-3. Check you're on the correct branch from PRD `branchName`. If not, check it out or create from main.
-4. Pick the **highest priority** user story where `passes: false`
-5. Implement that single user story
-6. Run quality checks (e.g., typecheck, lint, test - use whatever your project requires)
-7. Update CLAUDE.md files if you discover reusable patterns (see below)
-8. If checks pass, commit ALL changes with message: `feat: [Story ID] - [Story Title]`
-9. Update the PRD to set `passes: true` for the completed story
-10. Append your progress to `progress.txt`
-
-## Progress Report Format
-
-APPEND to progress.txt (never replace, always append):
-```
-## [Date/Time] - [Story ID]
-- What was implemented
-- Files changed
-- **Learnings for future iterations:**
-  - Patterns discovered (e.g., "this codebase uses X for Y")
-  - Gotchas encountered (e.g., "don't forget to update Z when changing W")
-  - Useful context (e.g., "the evaluation panel is in component X")
----
-```
-
-The learnings section is critical - it helps future iterations avoid repeating mistakes and understand the codebase better.
-
-## Consolidate Patterns
-
-If you discover a **reusable pattern** that future iterations should know, add it to the `## Codebase Patterns` section at the TOP of progress.txt (create it if it doesn't exist). This section should consolidate the most important learnings:
-
-```
-## Codebase Patterns
-- Example: Use `sql<number>` template for aggregations
-- Example: Always use `IF NOT EXISTS` for migrations
-- Example: Export types from actions.ts for UI components
-```
-
-Only add patterns that are **general and reusable**, not story-specific details.
-
-## Update CLAUDE.md Files
-
-Before committing, check if any edited files have learnings worth preserving in nearby CLAUDE.md files:
-
-1. **Identify directories with edited files** - Look at which directories you modified
-2. **Check for existing CLAUDE.md** - Look for CLAUDE.md in those directories or parent directories
-3. **Add valuable learnings** - If you discovered something future developers/agents should know:
-   - API patterns or conventions specific to that module
-   - Gotchas or non-obvious requirements
-   - Dependencies between files
-   - Testing approaches for that area
-   - Configuration or environment requirements
-
-**Examples of good CLAUDE.md additions:**
-- "When modifying X, also update Y to keep them in sync"
-- "This module uses pattern Z for all API calls"
-- "Tests require the dev server running on PORT 3000"
-- "Field names must match the template exactly"
-
-**Do NOT add:**
-- Story-specific implementation details
-- Temporary debugging notes
-- Information already in progress.txt
-
-Only update CLAUDE.md if you have **genuinely reusable knowledge** that would help future work in that directory.
-
-## Quality Requirements
-
-- ALL commits must pass your project's quality checks (typecheck, lint, test)
-- Do NOT commit broken code
-- Keep changes focused and minimal
-- Follow existing code patterns
-
-## Browser Testing (If Available)
-
-For any story that changes UI, verify it works in the browser if you have browser testing tools configured (e.g., via MCP):
-
-1. Navigate to the relevant page
-2. Verify the UI changes work as expected
-3. Take a screenshot if helpful for the progress log
-
-If no browser tools are available, note in your progress report that manual browser verification is needed.
-
-## Stop Condition
-
-After completing a user story, check if ALL stories have `passes: true`.
-
-If ALL stories are complete and passing, reply with:
-<promise>COMPLETE</promise>
-
-If there are still stories with `passes: false`, end your response normally (another iteration will pick up the next story).
-
-## Important
-
-- Work on ONE story per iteration
-- Commit frequently
-- Keep CI green
-- Read the Codebase Patterns section in progress.txt before starting

package/tasks/archive/2026-02-09-next-starter-dynamic-layers/prd.json

@@ -1,153 +0,0 @@
-{
-  "project": "basement-cli",
-  "branchName": "ralph/next-starter-dynamic-layers",
-  "description": "Replace 4 static templates with next-starter clone + dynamic technology layers to eliminate maintenance burden",
-  "userStories": [
-    {
-      "id": "US-001",
-      "title": "Create layer configuration system",
-      "description": "As a CLI maintainer, I want technology layers defined in a config file so that adding or modifying layers only requires changing one file.",
-      "acceptanceCriteria": [
-        "Add LAYER_CONFIG export to src/mergers/config.js alongside existing CMS_CONFIG",
-        "Each layer config defines: replaceFiles, additivePaths, dependencies, devDependencies",
-        "Add getLayerConfig(layer) helper function export",
-        "Config contains entries for webgl, webgpu, and experiment",
-        "webgpu is a deps-only layer (empty replaceFiles and additivePaths arrays)",
-        "webgl dependencies: @react-three/fiber ^9.5.0, @react-three/drei ^10.7.7, three ^0.182.0, @gsap/react ^2.0.0, @radix-ui/react-navigation-menu ^1.2.5, leva ^0.9.35, devDeps: @types/three ^0.182.0",
-        "webgpu dependencies: @react-three/fiber 10.0.0-alpha.2, @react-three/drei ^11.0.0-alpha.4, three ^0.182.0, leva ^0.9.35, lucide-react ^0.474.0, @radix-ui/react-navigation-menu ^1.2.5, devDeps: @types/three ^0.182.0",
-        "experiment dependencies: @react-three/fiber ^9.5.0, @react-three/drei ^10.7.7, three ^0.172.0, leva ^0.9.35, lucide-react ^0.474.0, @radix-ui/react-navigation-menu ^1.2.5, devDeps: @types/three ^0.182.0",
-        "Typecheck passes"
-      ],
-      "priority": 1,
-      "passes": true,
-      "notes": ""
-    },
-    {
-      "id": "US-002",
-      "title": "Extract layer files from templates",
-      "description": "As a CLI maintainer, I need the unique files from each template extracted into a layers/ directory so they can be overlaid on next-starter.",
-      "acceptanceCriteria": [
-        "Create layers/webgl/app/page.tsx copied from template/webgl/app/page.tsx (the one that imports DynamicCanvas)",
-        "Create layers/webgl/components/webgl/canvas/dynamic.tsx from template/webgl/components/webgl/canvas/dynamic.tsx",
-        "Create layers/webgl/components/webgl/canvas/index.tsx from template/webgl/components/webgl/canvas/index.tsx",
-        "Create layers/webgl/components/webgl/components/scene/index.tsx from template/webgl/components/webgl/components/scene/index.tsx",
-        "Create layers/experiment/components/layout/header/index.tsx from template/experiment/components/layout/header/index.tsx",
-        "Create layers/experiment/components/layout/navigation-menu.tsx from template/experiment/components/layout/navigation-menu.tsx",
-        "Create layers/experiment/lib/constats.ts from template/experiment/lib/constats.ts",
-        "Create layers/experiment/lib/utils/cn.ts from template/experiment/lib/utils/cn.ts",
-        "Create empty layers/webgpu/ directory with a .gitkeep file",
-        "All copied files are byte-for-byte identical to their template source files",
-        "Typecheck passes"
-      ],
-      "priority": 2,
-      "passes": true,
-      "notes": ""
-    },
-    {
-      "id": "US-003",
-      "title": "Implement layer injection function",
-      "description": "As the CLI system, I need to overlay layer-specific files on top of the next-starter base so that WebGL/WebGPU/Experiment projects get their unique components.",
-      "acceptanceCriteria": [
-        "Add injectLayer(targetDir, layer, spinner) function to src/mergers/index.js",
-        "Function resolves layer directory from CLI repo's layers/{type}/ path using __dirname",
-        "Function calls existing detectPathPrefix() to handle src/ directory structure",
-        "Function copies additive files from local layers/ directory (no tiged clone needed)",
-        "Function replaces files listed in replaceFiles config (overwrites base version with overwrite: true)",
-        "Function uses existing transformPath() for path prefix support",
-        "Function returns results object with replaced, copied, skipped, and failed arrays",
-        "When layer is 'default' or not found in config, function returns early with skipped result (no error)",
-        "Extend formatMergeResults() to display replaced entries with checkmark prefix",
-        "Typecheck passes"
-      ],
-      "priority": 3,
-      "passes": true,
-      "notes": "Reuses existing detectPathPrefix() and transformPath() from src/mergers/index.js. Layer files are local (no network request)."
-    },
-    {
-      "id": "US-004",
-      "title": "Change create.js to clone next-starter instead of template",
-      "description": "As a developer using the CLI, I want my project scaffolded from the latest next-starter repo so I always get the most up-to-date base.",
-      "acceptanceCriteria": [
-        "Change tiged source on line 124 of src/commands/create.js from 'github:basementstudio/basement-cli/template/${type}#${branch}' to 'github:basementstudio/next-starter#main'",
-        "Remove the BASEMENT_CLI_BRANCH env var usage for template (keep if used elsewhere)",
-        "Delete bun.lock file after clone if it exists (since dependencies will be modified)",
-        "Update download spinner text to say 'Downloading next-starter from GitHub...' instead of mentioning template type",
-        "Update error/troubleshooting message in catch block to reference basementstudio/next-starter instead of basementstudio/basement-cli/template/{type}",
-        "Typecheck passes"
-      ],
-      "priority": 4,
-      "passes": true,
-      "notes": ""
-    },
-    {
-      "id": "US-005",
-      "title": "Wire layer injection into create flow",
-      "description": "As the CLI system, I need the layer injection step between template download and CMS integration so layers are applied at the right time.",
-      "acceptanceCriteria": [
-        "Add new step between template download (step 3) and CMS integration (step 3.5) in src/commands/create.js",
-        "Layer injection only runs when type !== 'default'",
-        "Import injectLayer and formatMergeResults from src/mergers/index.js",
-        "Show ora spinner with layer name during injection (e.g. 'Applying webgl layer...')",
-        "Display injection results to user using formatMergeResults",
-        "On failure, show warning (not hard error) and continue with project creation",
-        "Typecheck passes"
-      ],
-      "priority": 5,
-      "passes": true,
-      "notes": "Must run BEFORE CMS integration because layers don't modify layout.tsx, and CMS mergers need the base layout intact."
-    },
-    {
-      "id": "US-006",
-      "title": "Config-driven dependency injection in package.json hydration",
-      "description": "As the CLI system, I need package.json hydration to read layer dependencies from LAYER_CONFIG instead of being hardcoded.",
-      "acceptanceCriteria": [
-        "Import getLayerConfig from src/mergers/config.js in create.js",
-        "Read layer dependencies from LAYER_CONFIG and merge into package.json: pkg.dependencies = { ...pkg.dependencies, ...layerConfig.dependencies }",
-        "Also merge devDependencies from layer config",
-        "CMS dependency injection (Sanity/BaseHub) still works unchanged",
-        "Animation library injection (GSAP/Framer Motion) still works unchanged",
-        "Simplify package.json handling: next-starter uses package.json directly (not package.template.json), remove the rename logic",
-        "Project name still set to user-provided name, version still set to 0.1.0",
-        "Typecheck passes"
-      ],
-      "priority": 6,
-      "passes": true,
-      "notes": "next-starter uses package.json not package.template.json, so the existsSync check for package.template.json can be simplified."
-    },
-    {
-      "id": "US-007",
-      "title": "Delete template directory",
-      "description": "As a CLI maintainer, I want the template/ directory removed so there's no duplicated code to maintain.",
-      "acceptanceCriteria": [
-        "Delete template/default/ directory and all contents",
-        "Delete template/webgl/ directory and all contents",
-        "Delete template/webgpu/ directory and all contents",
-        "Delete template/experiment/ directory and all contents",
-        "Grep source code for any remaining references to 'template/' paths and remove them",
-        "Grep source code for any remaining references to 'basementstudio/basement-cli/template/' and remove them",
-        "Typecheck passes"
-      ],
-      "priority": 7,
-      "passes": true,
-      "notes": ""
-    },
-    {
-      "id": "US-008",
-      "title": "End-to-end verification of all template types",
-      "description": "As a CLI maintainer, I need to verify all template + CMS + animation combinations produce working projects.",
-      "acceptanceCriteria": [
-        "Run npm link to install CLI locally",
-        "Run: basement -c test-default -d -no-cms -no-animation -claude -no-hooks — project creates successfully",
-        "Run: basement -c test-webgl -webgl -no-cms -no-animation -claude -no-hooks — project has R3F components in components/webgl/",
-        "Run: basement -c test-webgpu -webgpu -no-cms -no-animation -claude -no-hooks — project has WebGPU deps in package.json",
-        "Run: basement -c test-experiment -exp -no-cms -no-animation -claude -no-hooks — project has custom header with NavigationMenu",
-        "Run: basement -c test-sanity -webgl -sanity -gsap -claude -no-hooks — project has CMS + animation + layer deps",
-        "All generated projects pass bun install without errors",
-        "Typecheck passes"
-      ],
-      "priority": 8,
-      "passes": true,
-      "notes": "This is the final validation step. Each test project should be cleaned up after verification."
-    }
-  ]
-}

package/tasks/archive/2026-02-09-next-starter-dynamic-layers/progress.txt

@@ -1,115 +0,0 @@
-## Codebase Patterns
-- This is a plain JS (ESM) CLI project - no TypeScript, no typecheck command. "Typecheck passes" = no syntax errors.
-- Config pattern: export const CONFIG object + getXConfig(name) helper that returns config or null
-- src/mergers/config.js holds all configuration (CMS_CONFIG, LAYER_CONFIG)
-- src/mergers/index.js holds injection/merge logic (injectIntegration for CMS, injectLayer for layers)
-- src/commands/create.js is the main CLI flow
-- next-starter uses package.json directly (no template rename needed)
-- Layer files are local (in layers/ directory), CMS integration files are remote (fetched via tiged)
-- ESM requires __dirname workaround: `path.dirname(fileURLToPath(import.meta.url))`
-- CLI binary is basement (defined in package.json bin field)
-- Dependency injection order in create.js: layer deps → CMS deps → animation deps
-
----
-
-## 2026-02-09 - US-001
-- Implemented LAYER_CONFIG export with webgl, webgpu, experiment entries
-- Added getLayerConfig() helper function
-- Each layer config has: replaceFiles, additivePaths, dependencies, devDependencies
-- webgpu is deps-only (empty replaceFiles and additivePaths)
-- Files changed: src/mergers/config.js
-- **Learnings for future iterations:**
-  - The CLI project is plain JS ESM, verified by running `node --input-type=module -e "import {...}"`
-  - Config pattern follows existing CMS_CONFIG style - object + getter helper
-  - experiment uses three ^0.172.0 (not ^0.182.0 like webgl/webgpu) per PRD
----
-
-## 2026-02-09 - US-002
-- Created layers/webgl/ with app/page.tsx and 3 webgl component files from template/webgl/
-- Created layers/experiment/ with header, navigation-menu, constats.ts, cn.ts from template/experiment/
-- Created layers/webgpu/.gitkeep (empty layer, deps-only)
-- Verified all files are byte-for-byte identical using diff
-- Files changed: 9 new files in layers/
-- **Learnings for future iterations:**
-  - Layer files are local to CLI repo (no network clone needed, unlike CMS integrations)
-  - webgl is the only layer with a different app/page.tsx (imports DynamicCanvas)
-  - experiment has its own header and navigation-menu components
----
-
-## 2026-02-09 - US-003
-- Added injectLayer(targetDir, layer, spinner) function to src/mergers/index.js
-- Uses __dirname (via import.meta.url + fileURLToPath) to resolve layers/ directory
-- Reuses detectPathPrefix() and transformPath() for src/ directory support
-- Handles replaceFiles (overwrite: true) and additivePaths (overwrite: false) separately
-- Returns results with replaced, copied, skipped, failed arrays
-- Extended formatMergeResults() with defensive optional chaining for replaced/merged/copied/etc
-- Files changed: src/mergers/index.js
-- **Learnings for future iterations:**
-  - ESM doesn't have __dirname, must use: `const __dirname = path.dirname(fileURLToPath(import.meta.url))`
-  - Layer injection is purely local file copy (no network), unlike CMS integration which uses tiged
-  - formatMergeResults now handles both CMS results (merged) and layer results (replaced) gracefully
----
-
-## 2026-02-09 - US-004
-- Changed tiged source from basement/template/${type} to basementstudio/next-starter#main
-- Removed BASEMENT_CLI_BRANCH env var usage (was only used for template, not used elsewhere)
-- Added bun.lock deletion after clone (since deps will be modified by layers/CMS/animation)
-- Updated spinner text to "Downloading next-starter from GitHub..."
-- Updated troubleshooting message to reference basementstudio/next-starter
-- Kept dotenv import as it may be used by merger modules
-- Files changed: src/commands/create.js
-- **Learnings for future iterations:**
-  - dotenv is loaded in create.js but env vars are only used in merger modules (check-integration-merger.js)
-  - bun.lock must be deleted before modifying package.json deps, otherwise bun install may conflict
----
-
-## 2026-02-09 - US-005
-- Added layer injection step (3.2) between template download and CMS integration (3.5)
-- Layer injection only runs when type !== 'default'
-- Uses dynamic import for injectLayer and formatMergeResults from mergers/index.js
-- Shows ora spinner with layer name during injection
-- On failure, shows warning and continues (not a hard error)
-- Files changed: src/commands/create.js
-- **Learnings for future iterations:**
-  - Layer injection MUST happen before CMS integration to keep layout.tsx intact for CMS mergers
-  - Uses dynamic import (await import) consistent with existing CMS integration pattern
----
-
-## 2026-02-09 - US-006
-- Imported getLayerConfig from src/mergers/config.js at top of create.js
-- Layer deps injected from LAYER_CONFIG before CMS and animation deps
-- Simplified package.json handling: removed package.template.json logic (next-starter uses package.json directly)
-- CMS and animation dependency injection unchanged
-- Project name and version still set correctly
-- Files changed: src/commands/create.js
-- **Learnings for future iterations:**
-  - Dependency injection order matters: layer deps first, then CMS, then animation (later ones override earlier)
-  - next-starter uses package.json directly, no need for package.template.json rename logic
----
-
-## 2026-02-09 - US-007
-- Deleted template/default/, template/webgl/, template/webgpu/, template/experiment/ directories
-- Removed entire template/ directory (281 files, -27,778 lines)
-- Verified no remaining references to template/ paths in source code (src/)
-- Updated CLAUDE.md to reflect new architecture (layers/ instead of template/)
-- Files changed: template/ deleted, CLAUDE.md updated
-- **Learnings for future iterations:**
-  - CLAUDE.md should be kept in sync with architecture changes
-  - Only documentation files (CLAUDE.md, PRD) had remaining references to template/ paths, not source code
----
-
-## 2026-02-09 - US-008
-- Installed CLI locally via npm link
-- E2E test results (all passed):
-  - test-default: Created successfully, name=test-default, version=0.1.0, has next dep, no R3F deps
-  - test-webgl: Created with layer applied (replaced page.tsx, added 3 webgl components), R3F ^9.5.0, three ^0.182.0, @types/three ^0.182.0
-  - test-webgpu: Created with deps-only layer, R3F 10.0.0-alpha.2, drei ^11.0.0-alpha.4, lucide-react ^0.474.0
-  - test-experiment: Created with layer (header, navigation-menu, constats.ts, cn.ts), three ^0.172.0
-  - test-sanity: Combined webgl + sanity + gsap - all deps present (R3F, next-sanity, sanity, gsap, @gsap/react, sanity:extract script)
-- All 4 projects pass bun install without errors
-- CLI command is basement (not basement-starter as PRD mentions)
-- **Learnings for future iterations:**
-  - Skills installation may fail without network (bunx skills add), but it's non-blocking
-  - The sanity integration triggers an interactive Sanity setup prompt after project creation
-  - bun install correctly resolves all dependency versions including alpha packages
----