diff --git a/sentinel-agent/README.md b/sentinel-agent/README.md
new file mode 100644
index 0000000..5832b19
--- /dev/null
+++ b/sentinel-agent/README.md
@@ -0,0 +1,61 @@
+# Sentinel - Risk-Aware Validation & Escalation Agent
+
+A deterministic AI trust firewall built on Cortensor's decentralized inference network.
+
+## 🎯 Problem Solved
+
+High-stakes governance decisions require trust enforcement beyond simple AI outputs. Sentinel validates decisions through multi-tier redundancy and weighted consensus before execution.
+
+## 🏗️ Architecture
+
+- **Tier-based Redundancy**: 1→3→5 validator escalation
+- **Composite Trust**: 60% weighted agreement + 40% average confidence
+- **Thresholds**: Low (0.50), Balanced (0.65), Oracle (0.85)
+- **Cryptographic Artifacts**: SHA256 + signatures for audit trails
+
+## 🚀 Live Demo
+
+- **Frontend**: https://sentinel-rav-agent.vercel.app
+- **API Docs**: https://b524e8285ad3fbdb-139-59-8-118.serveousercontent.com/docs
+
+## 🛠️ Tech Stack
+
+- **Backend**: Python, FastAPI, PostgreSQL
+- **Frontend**: React, Tailwind CSS
+- **AI**: Cortensor (Sessions 78, 67, 79)
+
+## 📊 Cortensor Integration
+
+- Delegate (Session 78): Risk assessment
+- Completions (Session 78): AI inference
+- Validate (Sessions 67/79): Multi-tier validation with PoI/PoUW
+
+
+## 📖 Features
+
+✅ Multi-tier escalation (1→3→5 validators)
+✅ Weighted consensus computation
+✅ Cryptographic decision artifacts
+✅ PostgreSQL audit storage
+✅ Real-time Cortensor integration
+✅ Professional web interface
+
+## 🎥 Demo
+
+Visit https://sentinel-rav-agent.vercel.app and test with:
+- Navigate to "Agent" page
+- Enter governance decision
+- Watch real-time validation
+
+## 📜 License
+
+MIT License
+
+## 👤 Author
+- GitHub: [@0xZaid10]
+- Discord: zaid_678.
+```
+MIT License
+
+Copyright (c) 2025 
+