feat: Compiled 467-line Site Audit Report merging source code research…

- ".gsd/milestones/M018/slices/S01/SITE-AUDIT-REPORT.md"

GSD-Task: S01/T02
This commit is contained in:
jlightner 2026-04-03 19:49:14 +00:00
parent 35c9fad8ef
commit 31b72c731c
4 changed files with 558 additions and 1 deletions

View file

@ -17,7 +17,7 @@ Key constraint: The site is at http://ub01:8096. API base is /api/v1. Admin endp
- Estimate: 45m
- Files: .gsd/milestones/M018/slices/S01/AUDIT-FINDINGS.md, .gsd/milestones/M018/slices/S01/S01-RESEARCH.md
- Verify: test -f .gsd/milestones/M018/slices/S01/AUDIT-FINDINGS.md && grep -c '##' .gsd/milestones/M018/slices/S01/AUDIT-FINDINGS.md | grep -q '[3-9]'
- [ ] **T02: Compile Site Audit Report from research and audit findings** — Compile the final Site Audit Report by merging the S01-RESEARCH.md source code analysis with the T01 AUDIT-FINDINGS.md live verification results. The report is the primary deliverable of this slice — a single markdown document that Phase 2 planning references for all implementation decisions.
- [x] **T02: Compiled 467-line Site Audit Report merging source code research with live audit findings, documenting all 12 routes, 41 API endpoints, 13 data models, CSS architecture, and 8 Phase 2 integration risks** — Compile the final Site Audit Report by merging the S01-RESEARCH.md source code analysis with the T01 AUDIT-FINDINGS.md live verification results. The report is the primary deliverable of this slice — a single markdown document that Phase 2 planning references for all implementation decisions.
Steps:
1. Read S01-RESEARCH.md and AUDIT-FINDINGS.md

View file

@ -0,0 +1,467 @@
# Chrysopedia — Site Audit Report
**Audit Date:** 2026-04-03
**Site URL:** http://ub01:8096
**API Version:** v0.1.0 | **App Version:** v0.8.0
**Auditor:** GSD Auto-Mode (M018/S01)
---
## Executive Summary
Chrysopedia is a music production knowledge base that synthesizes technique articles from video transcripts using an LLM pipeline. The site runs as a self-hosted Docker Compose stack on `ub01` with 8 containers serving a React SPA frontend, FastAPI backend, Celery worker, PostgreSQL, Redis, Qdrant vector store, Ollama embedding service, and nginx reverse proxy.
**Scale at time of audit:** 83 technique pages, 2526 creators, 200 source videos, 7 topic categories. Single-admin tool with no authentication layer.
**Tech stack:**
- **Frontend:** React 18 + TypeScript + Vite (no UI library, no state management library, no CSS framework)
- **Backend:** FastAPI (async) + SQLAlchemy (async for API, sync for Celery) + Celery + PostgreSQL 16 + Redis 7 + Qdrant 1.13.2 + Ollama
- **Infrastructure:** Docker Compose on ub01 (172.32.0.0/24 subnet), nginx reverse proxy exposing port 8096
**Key architectural characteristics:**
- Zero external frontend dependencies beyond React, react-router-dom, and Vite
- Monolithic CSS (5,820 lines, single file, BEM naming, 77 custom properties)
- No authentication on any endpoint (admin routes are network-access-controlled)
- Dual SQLAlchemy strategy: async engine for FastAPI, sync engine for Celery pipeline
- Non-blocking pipeline side effects (embedding failures don't block page synthesis)
---
## Route Map
All 12 frontend routes verified in live browser session. No JavaScript blocking errors observed on any route.
| # | Route Pattern | Page Component | Auth | Render Status | Notes |
|---|---|---|---|---|---|
| 1 | `/` | Home | Public | ✅ OK | Hero search, 3-step process cards, popular topics, stats counters, Topics/Creators feature cards |
| 2 | `/search` | SearchResults | Public | ✅ OK | Sort (Relevance/Newest/Oldest/A-Z), tag/content match highlights, partial matches section |
| 3 | `/techniques/:slug` | TechniquePage | Public | ✅ OK | v2 body sections, ToC sidebar, citation superscripts, tag/plugin badges, key moments |
| 4 | `/creators` | CreatorsBrowse | Public | ✅ OK | Sort (Random/A-Z/Views), name filter, 12 genre filter buttons, creator cards with stats |
| 5 | `/creators/:slug` | CreatorDetail | Public | ✅ OK | Avatar, video count, genre breakdown, technique list. Known: duplicate genre badges from casing |
| 6 | `/topics` | TopicsBrowse | Public | ✅ OK | 7 category cards with icons, expandable sub-topic lists, topic filter input |
| 7 | `/topics/:category/:subtopic` | SubTopicPage | Public | ✅ OK | Breadcrumb nav, category badge, creator-grouped technique cards, sort dropdown |
| 8 | `/about` | About | Public | ✅ OK | Static: What Is, 6-step pipeline explanation, Who Maintains This |
| 9 | `/admin/reports` | AdminReports | Admin* | ✅ OK | Status filter tabs (All/Open/Acknowledged/Resolved/Dismissed), 0 reports at audit time |
| 10 | `/admin/pipeline` | AdminPipeline | Admin* | ✅ OK | 200 videos, status tabs, recent activity, worker status, debug toggle, per-video actions |
| 11 | `/admin/techniques` | AdminTechniquePages | Admin* | ✅ OK | 83 articles, 16 creators, table with filters/sort. Multi-source and creator filters |
| 12 | `*` (fallback) | → Redirect `/` | — | ✅ OK | SPA fallback returns 200 shell; react-router redirects unknown paths to `/` |
*Admin routes have no authentication gate — accessible to anyone with network access.
**Routing architecture:** Routes defined in a single `<Routes>` block in `App.tsx`. The nginx server block returns the SPA `index.html` for all paths, with react-router-dom v6 handling client-side routing. New routes are added to the `<Routes>` block — no other configuration required.
---
## Component Inventory
### Page Components (11)
| Component | Route | Key Features |
|---|---|---|
| Home | `/` | Hero search bar, useCountUp animation for stats, popular topic tags, nav cards |
| SearchResults | `/search` | Debounced search input, SortDropdown, result cards with highlights, partial match fallback |
| TechniquePage | `/techniques/:slug` | v1/v2 format-discriminated rendering (D024), TableOfContents sidebar, IntersectionObserver scroll-spy (D030), citation superscripts, key moments |
| CreatorsBrowse | `/creators` | Random default sort (D014), genre filter pills, name filter, CreatorAvatar |
| CreatorDetail | `/creators/:slug` | Avatar, genre breakdown badges, technique list with sort |
| TopicsBrowse | `/topics` | Category cards with colored left borders (D020), expandable sections (CSS grid 0fr/1fr animation), filter input |
| SubTopicPage | `/topics/:cat/:sub` | Breadcrumb navigation, creator-grouped cards, sort dropdown |
| About | `/about` | Static content, numbered pipeline step cards |
| AdminReports | `/admin/reports` | Status filter tabs, report list (empty at audit time) |
| AdminPipeline | `/admin/pipeline` | Video list with status tabs, worker status, debug toggle, per-video actions (Retrigger/Revoke/Re-run Stage) |
| AdminTechniquePages | `/admin/techniques` | Technique table with multi-source filter, creator filter, format badge, version count |
### Shared Components (11+)
| Component | Used By | Purpose |
|---|---|---|
| SearchAutocomplete | Nav bar, mobile panel | Global search with Ctrl+Shift+F shortcut. `globalShortcut` prop prevents duplicate handlers (KNOWLEDGE entry) |
| AdminDropdown | Nav bar | Hover-open at desktop (D028), tap-toggle on mobile |
| AppFooter | All pages | Version display, build date, GitHub link, About link |
| TableOfContents | TechniquePage | Sticky sidebar ToC with scroll-spy active indicator |
| SortDropdown | SearchResults, CreatorsBrowse, SubTopicPage | Reusable sort selector |
| TagList | TechniquePage, SearchResults | Renders tag/badge pills |
| CategoryIcons | TopicsBrowse | SVG icons per topic category |
| CreatorAvatar | CreatorsBrowse, CreatorDetail | Avatar display with fallback |
| CopyLinkButton | TechniquePage | Clipboard copy with tooltip |
| SocialIcons | CreatorDetail | Social media link icons |
| ReportIssueModal | TechniquePage | Content report submission form |
### Hooks (3)
| Hook | Purpose |
|---|---|
| useCountUp | Animated counter for homepage stats (technique/creator counts) |
| useSortPreference | Persists sort preference per page in localStorage |
| useDocumentTitle | Sets `<title>` per page — all 10 pages instrumented (D022) |
### Utilities (2)
| File | Purpose |
|---|---|
| citations.tsx | Citation superscript rendering and linking for technique page body sections |
| catSlug.ts | Category/subtopic slug generation and normalization |
### API Client
Single module `public-client.ts` (~600 lines) with typed `request<T>` helper. Relative `/api/v1` base URL (nginx proxies to API container). TypeScript interfaces for all response types defined in the same file. Vite dev proxy configured for local development (`/api``localhost:8001`).
**State management:** Local component state only (useState/useEffect). No Redux, Zustand, Context providers, or external state management.
---
## API Surface Map
### Public Endpoints (10)
| # | Method | Path | Response Shape | Verified | Notes |
|---|---|---|---|---|---|
| 1 | GET | `/health` | `{status, service, version, database}` | ✅ 200 | version="0.1.0", db=connected |
| 2 | GET | `/api/v1/stats` | `{technique_count, creator_count}` | ✅ 200 | 83 techniques, 26 creators |
| 3 | GET | `/api/v1/search?q=` | `{items, partial_matches, total, query, fallback_used}` | ✅ 200 | ⚠️ Uses `items` key, not `results`. Has keyword fallback (D009) |
| 4 | GET | `/api/v1/search/suggestions?q=` | `{suggestions: [{text, type}]}` | ✅ 200 | Technique name autocomplete |
| 5 | GET | `/api/v1/search/popular` | `{items: [{query, count}]}` | ✅ 200 | Popular searches from search_log (D025) |
| 6 | GET | `/api/v1/techniques?limit=&offset=` | `{items, total, offset, limit}` | ✅ 200 | Paginated, total=83 |
| 7 | GET | `/api/v1/techniques/random` | `{slug}` | ✅ 200 | Returns JSON slug, NOT a redirect |
| 8 | GET | `/api/v1/techniques/{slug}` | 22-field object (see below) | ✅ 200 | Rich response with eager-loaded relations |
| 9 | GET | `/api/v1/techniques/{slug}/versions` | `{items, total}` | ✅ 200 | Paginated version list |
| 10 | GET | `/api/v1/techniques/{slug}/versions/{n}` | Version detail | — | Router-confirmed, not probed |
**Technique detail fields (22):** title, slug, topic_category, topic_tags, summary, body_sections, body_sections_format, signal_chains, plugins, id, creator_id, creator_name, creator_slug, source_quality, view_count, key_moment_count, created_at, updated_at, key_moments, creator_info, related_links, version_count, source_videos.
### Browse Endpoints (5)
| # | Method | Path | Response Shape | Verified | Notes |
|---|---|---|---|---|---|
| 1 | GET | `/api/v1/creators?sort=&genre=` | `{items, total, offset, limit}` | ✅ 200 | sort=random\|alpha\|views, genre filter. total=25 (vs 26 in stats) |
| 2 | GET | `/api/v1/creators/{slug}` | 16-field object | ✅ 200 | Includes genre_breakdown, techniques list, social_links |
| 3 | GET | `/api/v1/topics` | `[{name, description, sub_topics}]` | ✅ 200 | ⚠️ Returns bare list, not paginated dict |
| 4 | GET | `/api/v1/topics/{cat}/{sub}` | `{items, total, offset, limit}` | ✅ 200 | Paginated technique list for subtopic |
| 5 | GET | `/api/v1/topics/{cat}` | `{items, total, offset, limit}` | ✅ 200 | Paginated technique list for category |
### Report Endpoints (3)
| # | Method | Path | Response Shape | Verified | Notes |
|---|---|---|---|---|---|
| 1 | POST | `/api/v1/reports` | Report object | ✅ 422 | Empty body → validation error (confirms endpoint exists) |
| 2 | GET | `/api/v1/admin/reports` | Report list | ✅ 200 | No auth required |
| 3 | PATCH | `/api/v1/admin/reports/{id}` | Updated report | — | Router-confirmed, not probed (destructive) |
### Pipeline Admin Endpoints (20+)
All under prefix `/api/v1/admin/pipeline/` (NOT `/api/v1/pipeline/`).
| # | Method | Path | Purpose | Verified |
|---|---|---|---|---|
| 1 | GET | `/admin/pipeline/videos` | Paginated video list with pipeline status | ✅ 200 |
| 2 | POST | `/admin/pipeline/trigger/{video_id}` | Trigger pipeline for video | Router-confirmed |
| 3 | POST | `/admin/pipeline/clean-retrigger/{video_id}` | Clean retrigger (wipe + reprocess) | Router-confirmed |
| 4 | POST | `/admin/pipeline/revoke/{video_id}` | Revoke active pipeline task | Router-confirmed |
| 5 | POST | `/admin/pipeline/rerun-stage/{video_id}` | Re-run specific pipeline stage | Router-confirmed |
| 6 | GET | `/admin/pipeline/events` | Pipeline event log | Router-confirmed |
| 7 | GET | `/admin/pipeline/runs` | Pipeline run history | Router-confirmed |
| 8 | GET | `/admin/pipeline/chunking-inspector/{video_id}` | Inspect chunking results | Router-confirmed |
| 9 | GET | `/admin/pipeline/embed-status` | Embedding/Qdrant health | Router-confirmed |
| 10 | GET | `/admin/pipeline/debug-mode` | Get debug mode state | ✅ 200 |
| 11 | POST | `/admin/pipeline/debug-mode` | Set debug mode state | Router-confirmed |
| 12 | GET | `/admin/pipeline/token-summary` | Token usage summary | Router-confirmed |
| 13 | GET | `/admin/pipeline/stale-pages` | Pages needing regeneration | ✅ 200 |
| 14 | POST | `/admin/pipeline/bulk-resynthesize` | Regenerate all technique pages | Router-confirmed |
| 15 | POST | `/admin/pipeline/wipe-all-output` | Delete all pipeline output | Router-confirmed |
| 16 | POST | `/admin/pipeline/optimize-prompt` | Trigger prompt optimization | Router-confirmed |
| 17 | POST | `/admin/pipeline/reindex-all` | Rebuild Qdrant index | Router-confirmed |
| 18 | GET | `/admin/pipeline/worker-status` | Celery worker health | ✅ 200 |
| 19 | GET | `/admin/pipeline/recent-activity` | Recent pipeline events | ✅ 200 |
| 20 | POST | `/admin/pipeline/creator-profile/{creator_id}` | Update creator profile | Router-confirmed |
| 21 | POST | `/admin/pipeline/avatar-fetch/{creator_id}` | Fetch creator avatar | Router-confirmed |
### Other Endpoints (2)
| # | Method | Path | Response Shape | Verified | Notes |
|---|---|---|---|---|---|
| 1 | POST | `/api/v1/ingest` | Ingest result | ✅ 422 | Transcript upload (empty body → validation) |
| 2 | GET | `/api/v1/videos` | `[{...}]` | ✅ 200 | ⚠️ Returns bare list, not paginated dict |
### Response Shape Inconsistencies
Three endpoints return bare lists instead of paginated dicts:
- `GET /api/v1/topics` → bare list
- `GET /api/v1/videos` → bare list
- Creator count mismatch: `/stats` says 26, `/creators` returns 25 (likely a 0-technique creator excluded from browse)
**New endpoints should follow the paginated `{items, total, offset, limit}` pattern.**
---
## Data Model
13 SQLAlchemy models in `backend/models.py`. Key relationships shown below.
### Core Content Models
| Model | Key Fields | Relationships |
|---|---|---|
| **Creator** | name, slug, genres, avatar_url, bio, social_links, featured | → SourceVideo (1:N), → TechniquePage (1:N) |
| **SourceVideo** | filename, youtube_url, folder_name, processing_status, pipeline_stage | → Creator (N:1), → TranscriptSegment (1:N), → KeyMoment (1:N) |
| **TranscriptSegment** | start_time, end_time, text | → SourceVideo (N:1) |
| **KeyMoment** | title, summary, start_time, end_time, topic_category, topic_tags | → SourceVideo (N:1) |
| **TechniquePage** | title, slug, summary, body_sections (JSONB), body_sections_format, signal_chains, plugins, topic_category, topic_tags, source_quality, view_count | → Creator (N:1), → Tag (M:N via technique_page_tags), → KeyMoment (M:N via TechniquePageVideo) |
| **TechniquePageVersion** | version_number, content_snapshot (JSONB), pipeline_metadata (JSONB) | → TechniquePage (N:1) |
### Supporting Models
| Model | Purpose |
|---|---|
| **RelatedTechniqueLink** | Directed link between technique pages (source → target with label) |
| **Tag** | Normalized tag with M:N join to TechniquePage |
| **TechniquePageVideo** | Join table: TechniquePage ↔ SourceVideo (multi-source pages) |
| **ContentReport** | User-submitted content reports with status workflow |
| **SearchLog** | Query logging for popular searches feature (D025) |
| **PipelineRun** | Pipeline execution tracking per video |
| **PipelineEvent** | Granular pipeline stage events for admin monitoring |
### Schema Notes
- No Alembic migrations directory — schema changes require manual DDL or adding Alembic setup
- `body_sections_format` discriminator enables v1/v2 format coexistence (D024)
- `topic_category` casing is inconsistent across records (known data quality issue — see KNOWLEDGE)
- Stage 4 classification data (topic_tags per moment) stored in Redis, not DB columns
---
## CSS Architecture
### Overview
| Property | Value |
|---|---|
| File | `frontend/src/App.css` |
| Lines | 5,820 |
| Unique classes | ~589 |
| Naming convention | BEM (`block__element--modifier`) |
| Theme | Dark-only (no light mode) |
| Custom properties | 77 in `:root` (D017) |
| Accent color | Cyan `#22d3ee` |
| Font stack | System fonts (no custom fonts) |
| Preprocessor | None |
| CSS Modules | None |
| Component library | None |
### Custom Property Categories (77 total)
The 77 CSS custom properties in `:root` cover these semantic categories:
- **Surface colors:** page background, card backgrounds, nav, footer, input backgrounds
- **Text colors:** primary, secondary, muted, inverse, link, heading
- **Accent colors:** primary cyan, hover/active states, focus rings
- **Badge colors:** Per-category badge pairs (background + text) for 7 topic categories
- **Status colors:** Success/warning/error/info for admin states
- **Border colors:** Default, hover, focus, divider
- **Shadow colors:** Elevation shadows, glow effects
- **Overlay colors:** Modal/dropdown overlays
### Breakpoints
Four distinct breakpoint values used across 10 `@media` blocks:
| Breakpoint | Usage |
|---|---|
| 480px | Narrow mobile — compact card layouts |
| 600px | Wider mobile — grid adjustments |
| 640px | Small tablet — content width changes |
| 768px | Desktop → mobile transition — sidebar collapse, layout shift |
**Inconsistency risk:** Four different mobile breakpoints means some components respond at different widths. Phase 2 should standardize to 23 breakpoints (e.g., 480px mobile, 768px tablet, 1024px desktop).
### Layout Patterns
- **Page max-width:** 64rem (widened from 48rem in M005, D019)
- **Technique page:** CSS grid 2-column (1fr main + 22rem sidebar), collapses at 768px
- **Card layouts:** CSS grid with `auto-fill, minmax(...)` for responsive grids
- **Collapsible sections:** `grid-template-rows: 0fr/1fr` animation pattern
- **Sticky elements:** ToC sidebar, reading header (via `position: sticky`)
### Dead CSS Risk
With 5,820 lines and ~589 classes in a single file, dead CSS accumulation is likely. No tooling (PurgeCSS, unused-css-detector) is configured. Phase 2 additions will increase this debt. **Recommendation:** Run a dead CSS analysis before starting Phase 2 frontend work, and consider splitting App.css by feature area or adopting CSS modules for new components.
---
## Infrastructure Stack
### Docker Services
| Service | Image | Container | Port | Volume |
|---|---|---|---|---|
| PostgreSQL 16 | postgres:16-alpine | chrysopedia-db | 5433:5432 | chrysopedia_postgres_data |
| Redis 7 | redis:7-alpine | chrysopedia-redis | 6379 (internal) | — |
| Qdrant 1.13.2 | qdrant/qdrant:v1.13.2 | chrysopedia-qdrant | 6333 (internal) | chrysopedia_qdrant_data |
| Ollama | ollama/ollama:latest | chrysopedia-ollama | 11434 (internal) | chrysopedia_ollama_data |
| API (FastAPI) | Dockerfile.api | chrysopedia-api | 8000 (internal) | Bind: backend/, prompts/ |
| Worker (Celery) | Dockerfile.api (same image) | chrysopedia-worker | — | Bind: backend/, prompts/ |
| Watcher | Dockerfile.api (same image) | chrysopedia-watcher | — | Bind: watch dir |
| Web (nginx) | Dockerfile.web | chrysopedia-web-8096 | 8096:80 | — |
### Network
- Compose subnet: 172.32.0.0/24 (D015)
- External access: nginx on port 8096 proxies to API container (port 8000 internal)
- Vite dev proxy: `/api``localhost:8001`
### Build Pipeline
- **Frontend:** Vite build producing static assets in `dist/`. Dockerfile.web runs `npm run build` then serves via nginx.
- **Backend:** Dockerfile.api installs Python dependencies, runs uvicorn (API) or celery (worker/watcher).
- **Build-time injection:** `__APP_VERSION__`, `__BUILD_DATE__`, `__GIT_COMMIT__` via Vite `define` (must use `JSON.stringify` — see KNOWLEDGE).
- **Docker ARG→ENV ordering:** ARG must precede ENV which must precede `RUN npm run build` (KNOWLEDGE entry).
### Deployment
```bash
# Standard deploy on ub01
cd /vmPool/r/repos/xpltdco/chrysopedia
git pull
docker compose build && docker compose up -d
```
No CI/CD pipeline. Manual git pull + compose rebuild. Alembic not configured — schema changes require manual DDL.
---
## Phase 2 Integration Recommendations
### Adding New Routes
1. Add `<Route path="/new-path" element={<NewPage />} />` to the `<Routes>` block in `App.tsx`
2. Create page component in `frontend/src/pages/NewPage.tsx`
3. Call `useDocumentTitle('Page Title — Chrysopedia')` in the component
4. No nginx configuration needed — SPA fallback handles all paths
5. **Risk:** No code splitting configured. All page components are bundled together. Phase 2 should add `React.lazy()` + `Suspense` for new heavy pages.
### Adding New API Endpoints
1. Create router file: `backend/routers/foo.py` with `APIRouter(prefix="/foo", tags=["foo"])`
2. Register in `backend/main.py`: `app.include_router(foo.router, prefix="/api/v1")`
3. Define Pydantic schemas in `backend/schemas.py`
4. Use paginated response pattern: `{items, total, offset, limit}` (not bare lists)
5. Admin endpoints go under `/api/v1/admin/` prefix
6. **Risk:** No request validation middleware, no rate limiting, no auth. Phase 2 should decide on auth strategy early (D033 defers Stripe Connect, but admin endpoints need protection at minimum).
### Adding New Frontend API Calls
1. Add TypeScript interface to `public-client.ts`
2. Add function using `request<T>` helper with appropriate HTTP method
3. Base URL is relative `/api/v1` — nginx proxies to backend
4. **Risk:** `public-client.ts` is already ~600 lines. Consider splitting by domain (search, techniques, creators, admin) in Phase 2.
### Adding New Database Models
1. Add model class to `backend/models.py`
2. Add Pydantic schema to `backend/schemas.py`
3. **No Alembic configured** — create migration manually or set up Alembic first
4. Use `datetime.now(timezone.utc).replace(tzinfo=None)` for timestamp defaults (D002)
5. Avoid column names that shadow ORM functions (e.g., `relationship`) — see KNOWLEDGE
6. **Risk:** All 13 models in one file. Phase 2 with significant new models should split `models.py` by domain.
### CSS Additions
1. Append to `frontend/src/App.css` using BEM naming convention
2. Use existing CSS custom properties for all colors — never hardcode hex values
3. Use existing breakpoint values (prefer 768px for the mobile/desktop split)
4. **Risk (highest):** Monolithic 5,820-line CSS file with no scoping. Any selector conflict affects the entire app. **Strong recommendation:** adopt CSS modules for all new Phase 2 components, and progressively migrate existing components during major refactors. At minimum, namespace new selectors (e.g., `.p2-chat__container` for Phase 2 chat feature).
### Key Phase 2 Risks
| # | Risk | Impact | Mitigation |
|---|---|---|---|
| 1 | Monolithic CSS (5,820 lines, no scoping) | Selector conflicts, dead CSS accumulation, maintenance cost | CSS modules for new components, PurgeCSS audit, progressive splitting |
| 2 | No authentication | Admin endpoints open to network, no user sessions for creator tools | Auth middleware decision needed before M019 (D033 defers Stripe) |
| 3 | No Alembic | Schema changes are manual DDL, no migration history, no rollback | Set up Alembic in M019 foundations milestone |
| 4 | No code splitting | Bundle grows with every new page, longer initial load | React.lazy + Suspense for new pages, route-based splitting |
| 5 | Single-file API client (~600 lines) | Merge conflicts, hard to find endpoints | Split by domain in M019 |
| 6 | Pipeline actively running | Schema/API changes can conflict with in-flight pipeline tasks | Freeze pipeline before major schema changes |
| 7 | All stale pages (83/83) | Bulk resynthesize triggered accidentally would overwrite all content | Add confirmation gate to bulk resynthesize endpoint |
| 8 | No CI/CD | Manual deploy, no automated testing in pipeline | Add GitHub Actions in M019 or M020 |
### Integration Points with Phase 2 Architecture (D031)
| Integration Point | Milestone | What Connects |
|---|---|---|
| INT-1: RAG Knowledge Graph | M020 (Core Experiences) | LightRAG (D032) hooks into existing Qdrant + adds NetworkX graph. New `/api/v1/chat` endpoint. |
| INT-2: Creator Portal Auth | M022 (Creator Tools) | Auth middleware (new) wraps existing admin + new creator endpoints. |
| INT-3: MinIO File Storage | M022 (Creator Tools) | MinIO (D035) serves signed URLs for presets/samples. New upload endpoints. |
| INT-4: Stripe Connect | Phase 3 (deferred) | Payment integration for creator tiers (D033 — demo-only in Phase 2). |
---
## Appendix: Audit Evidence
### Route Verification Summary
All 12 routes visited in a live browser session (desktop viewport 1280×800) at http://ub01:8096 on 2026-04-03. Every page rendered with consistent nav bar (logo, search, Home/Topics/Creators, Admin dropdown) and footer (version, build date, links).
**Visual observations:**
- Dark theme only — no light mode toggle
- No loading skeletons — pages render fully or show empty states
- Admin pages accessible without login (consistent with single-admin design)
- Minor 422 console warnings on admin pages (benign — from fetch of endpoints requiring parameters)
### Endpoint Probe Results
| Category | Total | Probed | Confirmed via Router | 200 OK | 422 (validation) | Not Probed |
|---|---|---|---|---|---|---|
| Public | 10 | 9 | 1 | 9 | 0 | 1 |
| Browse | 5 | 5 | 0 | 5 | 0 | 0 |
| Reports | 3 | 2 | 1 | 1 | 1 | 1 |
| Pipeline Admin | 21 | 5 | 16 | 5 | 0 | 16 |
| Other | 2 | 2 | 0 | 1 | 1 | 0 |
| **Total** | **41** | **23** | **18** | **21** | **2** | **18** |
18 endpoints not probed were confirmed from FastAPI router source code (POST/destructive operations not safe to execute during live audit with active pipeline).
### Data Snapshot at Audit Time
| Metric | Value |
|---|---|
| Technique pages | 83 |
| Creators (stats endpoint) | 26 |
| Creators (browse endpoint) | 25 |
| Source videos | 200 |
| Topic categories | 7 |
| Pipeline: queued | 148 |
| Pipeline: in progress | 3 |
| Pipeline: errored | 13 |
| Pipeline: complete | 36 |
| Stale pages | 83/83 |
| Celery workers | 1 online |
| Debug mode | Enabled |
| Content reports | 0 |
### Discrepancies Log
7 discrepancies found between source code research and live site verification:
| # | Area | Expected | Actual | Severity |
|---|---|---|---|---|
| 1 | Search response key | `results` | `items` + `partial_matches` + `fallback_used` | Medium |
| 2 | `/techniques/random` | HTTP redirect | JSON `{slug}` | Low |
| 3 | Creator count | 26 | 25 in browse (26 in stats) | Low |
| 4 | `/api/v1/topics` shape | Paginated dict | Bare list | Low |
| 5 | `/api/v1/videos` shape | Not documented | Bare list (not paginated) | Low |
| 6 | Pipeline admin prefix | `/api/v1/pipeline/` | `/api/v1/admin/pipeline/` | Medium |
| 7 | Creator genre badges | Clean categories | Duplicate badges from casing inconsistency | Low |
### Relevant Decision Register Entries
Key decisions from DECISIONS.md that affect Phase 2 integration:
- **D002:** Timestamp handling — `datetime.now(timezone.utc).replace(tzinfo=None)`
- **D009:** Async SearchService pattern for FastAPI, sync clients for Celery
- **D015:** Docker subnet 172.32.0.0/24
- **D017:** 77 CSS custom properties, cyan accent
- **D019:** 64rem max-width, 2-column grid layout
- **D024:** Format-discriminated rendering for v1/v2 body sections
- **D025:** Search query logging in PostgreSQL + Redis cache
- **D031:** Phase 2 milestone structure (M018M025)
- **D032:** LightRAG with Qdrant + NetworkX for chat/knowledge graph
- **D033:** Demo monetization — Stripe Connect deferred to Phase 3
- **D035:** MinIO for file/object storage

View file

@ -0,0 +1,16 @@
{
"schemaVersion": 1,
"taskId": "T01",
"unitId": "M018/S01/T01",
"timestamp": 1775245534335,
"passed": true,
"discoverySource": "task-plan",
"checks": [
{
"command": "test -f .gsd/milestones/M018/slices/S01/AUDIT-FINDINGS.md",
"exitCode": 0,
"durationMs": 8,
"verdict": "pass"
}
]
}

View file

@ -0,0 +1,74 @@
---
id: T02
parent: S01
milestone: M018
provides: []
requires: []
affects: []
key_files: [".gsd/milestones/M018/slices/S01/SITE-AUDIT-REPORT.md"]
key_decisions: []
patterns_established: []
drill_down_paths: []
observability_surfaces: []
duration: ""
verification_result: "File exists at 467 lines with 9 H2 sections — passes all three verification checks (file exists, ≥300 lines, ≥5 sections)."
completed_at: 2026-04-03T19:49:11.326Z
blocker_discovered: false
---
# T02: Compiled 467-line Site Audit Report merging source code research with live audit findings, documenting all 12 routes, 41 API endpoints, 13 data models, CSS architecture, and 8 Phase 2 integration risks
> Compiled 467-line Site Audit Report merging source code research with live audit findings, documenting all 12 routes, 41 API endpoints, 13 data models, CSS architecture, and 8 Phase 2 integration risks
## What Happened
---
id: T02
parent: S01
milestone: M018
key_files:
- .gsd/milestones/M018/slices/S01/SITE-AUDIT-REPORT.md
key_decisions:
- (none)
duration: ""
verification_result: passed
completed_at: 2026-04-03T19:49:11.326Z
blocker_discovered: false
---
# T02: Compiled 467-line Site Audit Report merging source code research with live audit findings, documenting all 12 routes, 41 API endpoints, 13 data models, CSS architecture, and 8 Phase 2 integration risks
**Compiled 467-line Site Audit Report merging source code research with live audit findings, documenting all 12 routes, 41 API endpoints, 13 data models, CSS architecture, and 8 Phase 2 integration risks**
## What Happened
Merged S01-RESEARCH.md (source code analysis) and AUDIT-FINDINGS.md (live site verification from T01) into a single comprehensive SITE-AUDIT-REPORT.md. The report has 9 major sections: Executive Summary, Route Map, Component Inventory, API Surface Map, Data Model, CSS Architecture, Infrastructure Stack, Phase 2 Integration Recommendations, and Appendix with audit evidence. Cross-referenced 12 DECISIONS.md entries where they affect Phase 2 integration. Identified 8 key Phase 2 risks with mitigations and documented 4 integration points from the Phase 2 milestone structure.
## Verification
File exists at 467 lines with 9 H2 sections — passes all three verification checks (file exists, ≥300 lines, ≥5 sections).
## Verification Evidence
| # | Command | Exit Code | Verdict | Duration |
|---|---------|-----------|---------|----------|
| 1 | `test -f .gsd/milestones/M018/slices/S01/SITE-AUDIT-REPORT.md && test $(wc -l < .gsd/milestones/M018/slices/S01/SITE-AUDIT-REPORT.md) -ge 300 && grep -c '^## ' .gsd/milestones/M018/slices/S01/SITE-AUDIT-REPORT.md | grep -q '[5-9]'` | 0 | ✅ pass | 100ms |
## Deviations
Research counted 43 endpoints; report reconciled to 41 after deduplication. No material deviation.
## Known Issues
None.
## Files Created/Modified
- `.gsd/milestones/M018/slices/S01/SITE-AUDIT-REPORT.md`
## Deviations
Research counted 43 endpoints; report reconciled to 41 after deduplication. No material deviation.
## Known Issues
None.