docs: rewrite user-facing documentation with screenshots (#11)

docs/index.md

@@ -2,7 +2,7 @@
 layout: home
 
 hero:
-  name: Compliance Scanner
+  name: Certifai
   text: AI-Powered Security Compliance
   tagline: Automated SAST, SBOM, DAST, CVE monitoring, and code intelligence for your repositories
   actions:
@@ -14,16 +14,16 @@ hero:
       link: /features/overview
 
 features:
-  - title: Static Analysis (SAST)
-    details: Automated security scanning with Semgrep, detecting vulnerabilities across multiple languages including OWASP patterns, GDPR issues, and OAuth misconfigurations.
+  - title: Smart Findings with AI Triage
+    details: Every finding is triaged by an LLM that considers severity, blast radius, and codebase context. You get a confidence score, rationale, and remediation guidance -- not just raw scanner output.
   - title: SBOM & License Compliance
-    details: Full software bill of materials with dependency inventory, vulnerability tracking, license compliance analysis, and export to CycloneDX/SPDX formats.
+    details: Full software bill of materials with dependency inventory, vulnerability tracking, license compliance analysis, and export to CycloneDX and SPDX formats.
   - title: Dynamic Testing (DAST)
     details: Black-box security testing of live web applications and APIs. Crawls endpoints, fuzzes parameters, and detects SQL injection, XSS, SSRF, and auth bypass vulnerabilities.
   - title: Code Knowledge Graph
-    details: Interactive visualization of your codebase structure. Understand function calls, class hierarchies, and module dependencies with community detection.
-  - title: Impact Analysis
-    details: When a vulnerability is found, see exactly which entry points and call chains are affected. Understand blast radius before prioritizing fixes.
+    details: Interactive visualization of your codebase structure. Understand function calls, class hierarchies, and module dependencies at a glance.
   - title: AI-Powered Chat
-    details: Ask questions about your codebase using RAG-powered AI. Code is embedded as vectors and retrieved contextually to give accurate, source-referenced answers.
+    details: Ask questions about your codebase using RAG-powered AI. Code is embedded and retrieved contextually to give accurate, source-referenced answers.
+  - title: MCP Integration
+    details: Expose your security data to LLM tools like Claude and Cursor through the Model Context Protocol. Query findings, SBOMs, and DAST results from any MCP-compatible client.
 ---