@sandrobuilds/tracerney 0.9.18 → 0.9.20

package/README.md

@@ -1,6 +1,8 @@
 # Tracerney
 
-Lightweight prompt injection detection for LLM applications. Runs 100% locally.
+Lightweight prompt injection detection for LLM applications. Runs 100% locally with zero data leaving your server.
+
+> 🚀 **Explore the full platform at [tracerney.com](https://www.tracerney.com)** — includes dashboard, analytics, API management, and team collaboration tools.
 
 ## Install
 
@@ -25,10 +27,10 @@ if (result.suspicious) {
 
 ## What's Included
 
-- **259 embedded attack patterns** — covers known injection techniques
-- **Local detection** — <5ms latency, zero network calls
+- **258 embedded attack patterns** — real-world injection techniques detected in real-time
+- **Local detection** — <5ms latency per prompt, zero network overhead
 - **Zero dependencies** — single npm package
-- **No data collection** — all detection happens in your process
+- **Privacy-first** — no data leaves your server, 100% local processing
 
 ## Result Object
 
@@ -64,18 +66,42 @@ if (result.suspicious) {
 
 ## Multi-Layer Runtime Defense
 
-**Layer 1:** Pattern-based detection with 259 embedded patterns
-- Fast local detection (<5ms)
+**Layer 1:** Pattern Matching (Always Free)
+- 258 real-world attack patterns in real-time
+- <5ms detection on modern hardware
 - Zero network overhead
-- Blocks known attack techniques
-- Marks suspicious prompts
-
-**Layer 2:** LLM Sentinel (Backend Verification)
-- Backend LLM verification for novel attacks
-- Context-aware analysis via OpenRouter
-- Returns structured threat metadata (class, fingerprint, confidence)
-- Rate limiting to prevent cost spikes
-- Only escalates suspicious Layer 1 matches
+- Local processing only
+- Detects: instruction overrides, role-play jailbreaks, context confusion, code execution risks, data extraction attempts, and more
+
+**Layer 2:** LLM Sentinel (Pro - $9/month)
+- **AI-powered response verification** — LLM-based analysis for novel attack patterns
+- **Context-aware scanning** — understands your application's specific security policies
+- **Delimiter salting** — prevents prompt injection through response boundaries
+- **Zero prompt storage** — responses are analyzed in-memory, never saved or logged
+- **Structured threat metadata** — detailed fingerprints for audit trails and tracking
+- **Advanced rate limiting** — prevents cost spikes with intelligent throttling
+
+## Pricing & Usage
+
+- **Free Tier:** 50 scans/month with Layer 1 pattern detection
+- **Pro Tier:** 2,500 scans/month with Layer 1 + Layer 2 LLM verification ($9/month)
+
+---
+
+## Ready for Advanced Protection?
+
+Layer 2 (LLM Sentinel) adds AI-powered verification with **context-aware** threat detection and **zero prompt storage** — all responses are analyzed in-memory and immediately discarded.
+
+**[Start Your Free Trial or Upgrade to Pro](https://www.tracerney.com/docs)** at tracerney.com
+
+Includes:
+- Dashboard with threat analytics
+- API key management
+- Team collaboration features
+- Detailed threat fingerprints for compliance
+- Priority support for Pro members
+
+---
 
 ## License
 

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@sandrobuilds/tracerney",
-  "version": "0.9.18",
-  "description": "Transparent proxy runtime sentinel for prompt injection defense",
+  "version": "0.9.20",
+  "description": "Lightweight prompt injection detection with Layer 1 (258 patterns) + Layer 2 (AI verification). Runs locally with zero data storage. Upgrade to Pro for context-aware threat analysis at tracerney.com",
   "type": "module",
   "main": "dist/index.js",
   "types": "dist/index.d.ts",