Wecontent

Content Automation System (CAS) — an internal tool prototype for FMCG beauty/personal care brand operations in Thailand.

Built by JLC GROUP CO.,LTD · License: MIT

---

What this is

Wecontent is an internal operations platform that takes a marketing team from brief → script → clip → publish → performance tracking in one place. The prototype unifies four workflow phases that today live in separate tools (Google Docs, Premiere, Shopee/Lazada seller centres, spreadsheets).

Four phases the system covers

1. Content Creation — SKU-aware script generator with 4 modular phases (customize → generate → review → export), Google Slides / Drive handoff, multi-platform presets (TikTok / Shopee / Lazada / IG / FB).
2. Script & Clip Review — Upload script + video to Drive, AI reads and returns severity-tagged notes (MUST-FIX / NICE / POLISH), per-note comment + regen loop, approval gate from script → clip.
3. Performance Tracking — Cross-platform ads + organic metrics (Meta / TikTok / Google / Shopee / Lazada), KOL attribution, creator ROI.
4. Reporting — Auto-generated weekly / monthly briefs, exec summary, competitor moves, customer voice aggregation.

---

Quick start

The prototype is a single-file HTML app — no build step, no server, no database. Open and use.

# Clone
git clone https://github.com/jlc-group/Wecontent.git
cd Wecontent

# Open prototype in your browser
open prototype/dashboard.html       # macOS
xdg-open prototype/dashboard.html   # Linux
start prototype/dashboard.html      # Windows

Current prototype stack: vanilla HTML + Alpine.js 3.13.5 + Tailwind CDN + Lucide icons. Everything runs client-side. All data is mock for demo purposes.

---

Folder map

Wecontent/
├── README.md                 ← you are here
├── LICENSE                   ← MIT
├── DOCS_INDEX.md             ← narrative overview of all spec docs
├── CLAUDE.md                 ← working notes for AI pair-programming
│
├── 00_Master_PRD.md          ← product requirements document
├── 01_System_Architecture.md
├── 02_Data_Model.md
├── 03_User_Roles_Permissions.md
├── 04_Phase1_Content_Creation.md
├── 05_Phase2_Video_Review.md
├── 06_Phase3_Performance_Tracking.md
├── 07_Phase4_Reporting.md
├── 08_Settings_Admin.md
├── 09_AI_Prompt_Library.md
├── 10_Team_SOP_RACI.md
├── 11_Integration_Specs.md
├── 12_Tech_Stack_Recommendation.md
├── 13_Phase_Rollout_Plan.md
├── 14_Ads_Data_Contract.md
├── 15_Data_Sources_Spec.md
├── MASTER_ARCHITECTURE.md    ← system-wide diagram + data flow
│
├── prototype/
│   ├── dashboard.html        ← main interactive prototype (v0.3.8.1, ~7000 lines)
│   └── dashboard_mockup.html ← earlier static mockup
│
├── deploy-notes/
│   ├── CHANGELOG.md          ← version history
│   ├── contracts/            ← data contracts between modules
│   ├── learning/             ← decisions + retrospectives
│   └── templates/            ← reusable boilerplate
│
├── templates/
│   ├── cursor_rules.md              ← coding rules for AI editors
│   ├── performance_tracker_schema.md
│   ├── script_template_teams.md     ← script format per platform
│   └── sku_brand_seed.md            ← mock SKU + brand seed data
│
├── skills/                   ← JLC custom skills (see skills/README.md)
│   ├── project/              ← 9 CAS-specific skills (cas-script-generator, cas-video-qa, fmcg-claim-library, ...)
│   ├── library/              ← 27 skill folders (branding, creative, kol-management, ...)
│   └── bundles/              ← packaged .skill archives
│
└── plugins/                  ← marketplace plugins (see plugins/README.md)
    ├── brand-voice/          ← from knowledge-work-plugins marketplace
    ├── data/  marketing/  engineering/  finance/
    ├── legal/  human-resources/  customer-support/
    ├── operations/  product-management/  productivity/
    └── marketplace-manifest.json

---

Tech stack (current + planned)

Layer	Current (prototype)	Planned (production)
Frontend	HTML + Alpine.js + Tailwind CDN	Next.js 14 + TypeScript + shadcn/ui
Backend	— (all mock)	Node.js + tRPC or FastAPI
DB	—	PostgreSQL + Prisma
Queue	—	BullMQ (Redis)
AI	mock responses	GPT-4 / Claude / Gemini + local whisper for video
Storage	—	Google Drive + S3
Auth	mock	Google OAuth + RBAC (see 03_User_Roles_Permissions.md)
Deployment	open-file	Vercel + Railway / AWS ECS

See 12_Tech_Stack_Recommendation.md for the rationale.

---

Roadmap

v0.3.x — current prototype iteration (interactive HTML, UX validation)

v0.4.x — data layer (real schemas from 02_Data_Model.md, Google Drive integration)

v0.5.x — AI pipeline (real script generation + video review behind feature flag)

v1.0 — production (auth, multi-tenant, ads data ingestion live)

See 13_Phase_Rollout_Plan.md for the full phased plan.

---

Version history

Highlights from deploy-notes/CHANGELOG.md:

Version	Date	Highlight
v0.3.8.1	2026-04-21	Renamed menu to "รีวิวคลิป + สคริปต์"
v0.3.8	2026-04-21	Script & Clip dual-mode review, workflow gating
v0.3.7	2026-04-21	Video Review: Drive upload + reasoning + per-note regen
v0.3.6	2026-04-20	Vanilla JS dropdown (fixed Alpine @click.away bug)
v0.3.5	2026-04-20	KOL Cockpit refactor (3 zones, 8 blocks)
v0.3.4	2026-04-19	Campaign tab in Create module, customize section

Full history → deploy-notes/CHANGELOG.md

---

Skills + plugins

Two complementary AI capability folders ship with this repo:

• skills/ — 27 JLC-authored skills covering Thai FMCG beauty workflows (branding, KOL, trade marketing, regulatory, forecasting, NPD, live commerce, etc.) plus packaged .skill bundles.
• plugins/ — 11 vendored copies of plugins from the public knowledge-work-plugins marketplace (brand-voice, data, marketing, engineering, finance, legal, HR, customer-support, operations, product-management, productivity).

Skills are our domain knowledge. Plugins are general-purpose, third-party, and best installed from the marketplace for live updates.

---

Contributing

This is an internal JLC Group project, but we accept PRs for:

• Bug reports on the prototype (open an issue with browser + steps to reproduce)
• UX / copy suggestions for Thai beauty-industry terminology
• Prompt library improvements (09_AI_Prompt_Library.md)
• Data contract refinements (14_Ads_Data_Contract.md, 15_Data_Sources_Spec.md)
• New or improved skills under skills/library/ (use the skill-creator skill to scaffold)

Branch strategy:

• main — stable, tagged releases
• develop — integration branch
• feature/<name> — feature branches off develop

---

License

MIT — see LICENSE. Built and maintained by JLC Group.

All brand names, SKU codes, KOL handles, and numeric examples in this repo are mock data for prototyping only and do not represent real commercial agreements.