◈ EngageChain

AI-Native Opinion Protocol — powered by GenLayer Intelligent Contracts

Submit any opinion or proposal. AI validators evaluate it through multi-validator consensus, and the result is stored permanently on-chain. You choose whether GenLayer evaluates it or you bring your own AI.

---

What It Does

Submit opinion → Choose AI validator → Consensus → Verdict stored on-chain

1. Submit — rich text editor, up to 2000 characters, signed on-chain
2. Choose validation — GenLayer LLM consensus or your own AI analysis
3. Evaluate — AI produces structured analysis (summary, sentiment, key points, recommendation)
4. Finalize — verdict stored permanently with a unique on-chain ID
5. Lookup — anyone can query any opinion by ID, result shown directly in the page

---

Run Locally

Zero dependencies. Just Node.js ≥ 14.

git clone https://github.com/Said1235/Engagechain.git
cd Engagechain
node server.js
# Open http://localhost:3001

server.js serves the frontend and proxies /api/rpc to studionet — all features work locally including submit, evaluate, and finalize.

---

Deploy to Vercel

# Option A — Vercel CLI
vercel deploy

# Option B — GitHub import
# Push to GitHub → vercel.com → New Project → import repo → Deploy

No environment variables needed. Vercel auto-detects api/rpc.js as a serverless function.

---

Features (v3)

Feature	Detail
🖊️ Rich text editor	Bold, italic, underline, lists — contenteditable with toolbar
🤖 Dual AI validation	GenLayer LLM consensus or paste your own AI analysis
📊 Consensus card	Emoji result card: Consensus / Decision / Confidence
🔍 In-page lookup	Opinion data displayed directly in the page, not just terminal
💾 Session persistence	Wallet stays connected after page refresh (localStorage)
⚡ Silent stats	Total submissions loads quietly without terminal noise
📺 Resizable terminal	Drag the bottom handle to resize; toggle to collapse

---

Contract

Address: 0x7D361211C5C2cB0A6066D583527F45EaE544e468
Network: GenLayer Studionet
View in Studio: studio.genlayer.com

Methods

Method	Type	Args	Returns
submit_opinion	write	text: str	opinion ID
evaluate_opinion	write	opinion_id: str	AI analysis object
submit_with_external_ai	write	opinion_id: str, analysis: str	normalized AI object
finalize_opinion	write	opinion_id: str, verdict: str	{id, status}
get_total_submissions	read	—	total count string
get_resolution_data	read	opinion_id: str	full opinion object
get_status	read	opinion_id: str	status string
get_all_opinions	read	—	all opinions dict

---

How AI Evaluation Works

def get_analysis() -> str:
    raw = gl.nondet.exec_prompt(task)          # LLM call — non-deterministic block
    parsed = json.loads(raw)
    return json.dumps(parsed, sort_keys=True)  # deterministic for strict_eq

result_str  = gl.eq_principle.strict_eq(get_analysis)  # validators agree byte-for-byte
result_json = json.loads(result_str)                    # deserialize OUTSIDE nondet

Multiple validators independently run the same prompt. strict_eq confirms they all produced equivalent results. No single party controls the output.

---

File Structure

api/
  rpc.js           Vercel serverless proxy → studionet (dual-port 443/8443)
contract/
  engagechain.py   GenLayer Intelligent Contract (Python)
app.js             Frontend logic
index.html         UI
styles.css         Styles
server.js          Local dev server (no npm install)
favicon.svg        EngageChain hexagon icon
vercel.json        maxDuration: 60s for AI consensus calls
package.json       Node 24.x engine pin

---

Stack

Smart Contract	Python · py-genlayer:test
Network	GenLayer Studionet
Frontend	Vanilla JS · HTML · CSS · Orbitron + Inter + Space Mono
Proxy	Node.js https module · dual-port retry (443 → 8443)
Deploy	Vercel Serverless Functions

---

Built on GenLayer · by @Said1235